United States*
         Environmental Protection
         Agency
Policy, Planning,
And Evaluation
(2122)
EPA 230-R-95-002
June 1995
         Climate Change Mitigation
         Strategies In The Forest
         And Agriculture Sectors
               !l
                                 'I
r z. -

-------

-------
 CLIMATE CHANGE MITIGATION STRATEGIES
IN THE FOREST AND AGRICULTURE SECTORS
        U.S. ENVIRONMENTAL PROTECTION AGENCY
       OFFICE OF POLICY, PLANNING, AND EVALUATION
              CLIMATE CHANGE DIVISION
              WASHINGTON, D.C., U.S.A.
                   JUNE 1995

-------

-------
                              ACKNOWLEDGEMENTS
This report is the result of an extensive collaborative effort by a number of organizations and
individuals. The work was sponsored and coordinated by the EPA Office of Policy, Planning,
and Evaluation (OPPE), in Washington, DC. Richard Morgenstern, Maryann Froehlich, Dennis
Tirpak, and Jane Leggett-Emil provided important guidance and support for the project.
Kenneth Andrasko, William Hohenstein, and Steven Winnett provided technical guidance,
coordination, and review throughout the project.

A number of individuals contributed to the design and coordination of the research,
including: Thomas Murphy, Peter Beedlow, and Robert Dixon at the EPA Environmental
Research Laboratory in Corvallis, OR; Rosemarie Russo, Lee Mulkey, and Tom Barnwell at
the EPA Environmental Research Laboratory in Athens, GA; Courtney Riordan at the EPA
Office of  Research and Development; Andy Manale at the Water and Agricultural Policy
Division of OPPE; Thomas Hamilton, Fred Kaiser, William Sommers and David Darr at the
USDA Forest Service Research Office; Richard Haynes, John Mills, Judy Mikowski, and Del
Thompson at the USDA Forest Service Pacific Northwest Research Station; Peter Ince and
Ken Skog at the USDA Forest Service Forest Products Laboratory, Madison, Wl; Carol
Whitman and Gary Evans at the USDA Global Change Office; and Kate Heaton at the Bruce
Company.

Studies for this analysis were conducted by: David Turner, Jeffrey Lee, Greg Koerper, and
Jerry Barker at the EPA Environmental Research Laboratory in Corvallis, OR; Richard
Haynes,  Ralph Alig, and Eric Moore at the USDA Forest Service Pacific Northwest Research
Station in Portland, OR; Richard Birdsey and Linda Heath at the USDA Forest Service in
Washington DC;  John Perez-Garcia at the Center for International Trade in Forest Products,
University of Washington, Seattle, WA; Aziz Bouzaher, Derald Holtkamp, Randall Reese, and
Jason Shogren at the Center for Agricultural and Rural Development, Iowa State University,
Ames, IA; Anthony Donigian, Avinash Patwardhan,  and Radha Chinnaswamy at AQUA
TERRA Consultants, Mountain View, CA; Tom Barnwell and Robert Jackson at the EPA
Environmental Research  Laboratory in Athens, GA; Kevin Weinrich and Allen Rowell at
Computer Sciences Corporation, Athens, GA; Vern Cole at the Natural Resources Ecology
Laboratory, Fort Collins, CO; Bruce McCarl at Texas A&M University, College Station, TX;
John (Mac) Callaway at RCG/Hagler Bailly, Boulder, CO; and Alice Cialella and Changsheng
Li at the  Bruce Company, Washington, DC.

Assistance in preparing this report synthesizing the study results was provided by Frances
Sussman, David Reiner, Jeff Fiedler, Doug Keinath, Susan Barvenik, and Roger Schwabacher
at ICF Incorporated, Washington, DC.

-------

-------
                                 TABLE OF CONTENTS
Executive Summary	ES-1


1.      Introduction  	   "I


2.      Background: The Carbon Cycle, Historic Land Use Change, and Mitigation
       Strategies	   3
       2.1     The Global Carbon Cycle and Carbon Sector Budgets	   4
       2.2     Historic Land Use Change: U.S. Forests	   6
       2.3     Historic Land Use Change:  Agriculture  	   8
       2.4     Overview of Scenarios	   8


3.      Methodology	   8
       3.1     Forest Sector Models	  12
       3.2     Agriculture Sector Models 	  13


4.      Tree Planting Scenarios 	  15
       4.1     Tree Planting on Marginal Cropland and Pastureland	  16
       4.2     Conservation Reserve and Wetlands Reserve Programs	  23
       4.3     The Agriculture Sector: Implications for Federally Funded Tree Planting  	  29


5.     Other  Forest Policy Scenarios	  35
       5.1     Increased Rates of Recycled Paper Utilization	  35
       5.2     Reduced National Forest Harvests	  38
       5.3     Increased Use of Biomass Energy  	  41
       5.4     Effectiveness of Combined Scenarios	  46


6.     Modified Agricultural Practices  	  51
       6.1     Modified Tillage Practices	  52
       6.2     Winter Cover Crops	  58


7.     Conclusions: Directions for Future Research 	  60
 References
61

-------

-------
                                    LIST OF FIGURES
2.      Background: The Carbon Cycle, Historic Land Use Change, and Mitigation Strategies

Figure 2-1  The Global Carbon Cycle	,	4

Figure 2-2 The Forest Sector Carbon Budget  	5

Figure 2-3 Carbon Storage on U.S. Timberland, by Region	7


3.      Methodology

Figure 3-1  Forest Sector Models	11

Figure 3-2 Agriculture Sector Models	11


4.      Tree Planting Scenarios

Figure 4-1  Carbon Storage in U.S. Forest Ecosystems by Forest Ecosystem Component	15

Figure 4-2 Geographical Regions Corresponding to Enrollment Scenarios	18

Figure 4-3 Total Carbon Storage on U.S. Public and Private Timberland: Range of Model
Results for Base Case Scenario	19

Figure 4-4 Total Carbon Storage on U.S. Public and Private Timberland: Increases
Relative to Base Case for Tree Planting Scenarios  	20

Figure 4-5 U.S. Softwood Inventory Projections for Privately Owned Timberland: Base
Case and Tree Planting Scenarios	22

Figure 4-6 Annual Acreage Under CRP Scenarios	25

Figure 4-7 Composition of Reserve Scenarios	26

Figure 4-8 Carbon Storage Under Reseive Program Scenarios: Increases Relative
to Base Case	27

Figure 4-9 Impacts on Agriculture for CRP and WRP Scenarios, Change from Base Case  	30

Figure 4-10  Average Annual  Net Carbon Accumulation: Tree Planting On Three Land Types	32
5.
Other Forest Policy Scenarios
Figure 5-1 Total Carbon Storage on U.S. Public and Private Timberland: Increases
Relative to Base Case for the Increased Recycling Scenario	
Figure 5-2  U.S. Softwood and Hardwood Inventory Projections for Privately Owned
Timberland: Base Case and Increased Recycling Scenario	
                                                                                 36
                                                                                 38

-------
Figure 5-3  U.S. Private Timber Inventory: Changes Relative to Base Case for National
Forest Harvest Scenario	41

Figure 5-4  Projected Global Production Increases in Response to Habitat Preservation,
by Region: 1995	42

Figure 5-5  Total Carbon Storage on U.S. Public and Private Timberland: Changes
Relative to Base Case for Increasing Biomass Energy Scenario  	44

Figure 5-6  U.S. Private Timber Inventory: Changes Relative to Base Case for Increasing
Biomass Energy Scenario  	45

Figure 5-7  Total Carbon Storage on U.S. Public and Private Timberland: Changes Relative
to Base Case for Combined Scenarios	48

Figure 5-8  U.S. Private Softwood Timber  Inventory: Changes  Relative to Base Case
for Combined Scenarios	49

Figure 5-9  U.S. Private Hardwood Timber Inventory:  Changes Relative to Base Case
for Combined Scenarios	49


6.     Modified Agricultural Practices

Figure 6-1   RAMS Study Region	53

Figure 6-2  Tillage Distributions for Alternative Scenarios  	54

Figure 6-3  Major Rotations: Base Case and Targeted Levels of Winter Cover Crops	58
                                             in

-------
                                     LIST OF TABLES
Executive Summary

Table ES-1 Annual Carbon Accumulation for the Forest and Agriculture Sector Scenarios,
Decadal Intervals	
                                                                                ES-5
4.     Tree Planting Scenarios

Table 4-1  Average Storage of Carbon in Forest Ecosystems by U.S. Region 	16

Table 4-2  Regional Distribution of New Forest Under Alternative Enrollment Schedules  . . .	18

Table 4-3  Annual Carbon Accumulation on U.S. Public and Private Timberland Relative
to Base Case for Three Tree Planting Scenarios, Decadal Intervals  	21

Table 4-4  U.S. Softwood Harvests for Tree Planting Scenarios: All Owners	22

Table 4-5  Stumpage Prices in the Base Case	.23

Table 4-6  Stumpage Prices in the Tree Planting Scenarios  	24

Table 4-7  Annual Carbon Accumulation for CRP and WRP Scenarios Relative to Base Case,
Decadal Intervals	27

Table 4-8  Impacts of Reserve Program Scenarios on Agricultural Production and Producer
Prices: Percent Change Relative to Base Case, Decadal Intervals	28

Table 4-9  Impacts of Tree Planting on Different Land Bases	32

Table 4-10 Impacts of Tree Planting on Cropland with Varying Enrollment Strategies	33

Table 4-11 Economic Welfare Effects of Federal Subsidies for Tree Planting on Cropland 	34

Table 4-12 Price Effects of Federal Subsidies for Tree Planting on Cropland:
Fisher Price Indices	34
 5.
Other Forest Policy Scenarios
Table 5-1 Annual Carbon Accumulation on U.S. Public and Private Timberland Relative
to Base Case for Increased Recycling Scenario, Decadal Intervals	37

Table 5-2 Changes in Harvest Levels for the Increased Recycling Scenario,
Relative to the Base Case   	37

Table 5-3 Stumpage Prices in the Increased Recycling Scenario	39

Table 5-4 National Forest Harvest Under Base Case and Reduced National Forest
Harvest Scenario	39

Table 5-5 Base Case Harvest Projections by Region: Public and Private Timberland Owners	40
                                              iv

-------
Table 5-6  Stumpage Prices in the Reduced National Forest Harvest Scenario 	42

Table 5-7  Biomass Fuelwood Consumption: Assumptions for Base Case and Biofuel Scenario,
Decadal Intervals	43

Table 5-8  Annual Carbon Accumulation on U.S. Public and Private Timberland Relative
to Base Case for Increased Use of Biomass Energy, Decadal Intervals	46

Table 5-9  Stumpage Prices in the Increased Biomass Energy Scenario	46

Table 5-10 Combination Scenarios  	47

Table 5-11  Annual  Carbon Accumulation on U.S. Public and Private Timberland Relative
to Base Case for Combined Scenarios, Decadal Intervals	50

Table 5-12 Stumpage Prices in the Combination Scenarios	51
6.     Modified Agricultural Practices

Table 6-1  Soil Organic Carbon Accumulation for the Conservation Tillage Scenarios
in the RAMS Study Region   	54

Table 6-2  Impact of Short Term Yield Assumptions on the Introduction of Conservation Tillage  ... 56

Table 6-3  Impact of Short Term Yield Assumptions on Major Crop Rotations	56

Table 6-4  Regional Variation in Soil Carbon Accumulation	57

Table 6-5  Soil Organic Carbon Accumulation for the Winter Cover Crop Scenario  	59

-------
                                     LIST OF BOXES


1.      Introduction

Box 1-1  Actions to Mitigate Climate Change  	1

Box 1 -2  Studies Conducted for this Assessment	3


2.      Background: The Carbon Cycle, Historic Land Use Change, and Mitigation Strategies

Box 2-1  Definition of Timberland and Woodland 	6

Box 2-2  Summary of Scenarios	9


3.      Methodology

Box 3-1  Summary of Key Models Used for this Assessment	10

Box 3-2  Estimating  Forest Carbon: FCM and FORCARB  	13


4.      Tree Planting Scenarios

Box 4-1  Investment Behavior in TAMM/ATLAS	17


5.      Other Forest Policy Scenarios

Box 5-1  Paper Recycling in the U.S	35


6.      Modified Agricultural Practices

Box 6-1  The Effects of Crop Rotation, Tillage Practice, and Climate on Soil Carbon
for Two Neighboring Climate Divisions	55
                                            VI

-------

-------
                                   EXECUTIVE SUMMARY
Human activities, particularly the use of fossil
fuels for energy, increase atmospheric levels of
greenhouse gases, which may induce changes
in the earth's climate over the coming
decades. While there remains considerable
uncertainty about the rate and magnitude of
possible climatic changes, the Intergovern-
mental Panel on Climate Change, under the
auspices of the World Meteorological
Organization and the United Nations
Environmental Program, estimates that rising
concentrations of greenhouse gases are likely
to increase global temperatures by between
1.5°C and 4.5°C (2.7°F and 8.1°F) over the next
century (IPCC  1992).

Together with temperature rise, related climatic
changes, including rising sea level and altered
storm patterns, are likely to  have adverse
impacts, such  as loss of wetlands and coastal
property, changes in water supply, and
reductions in agricultural productivity in some
areas. The U.S. Government is examining
policies that may help reduce atmospheric
concentrations of greenhouse gases, the most
important of which is carbon dioxide (CO2).

Because vegetation and soil contain about
three times as much carbon as the
atmosphere, terrestrial ecosystems offer an
opportunity to absorb and store (sequester) a
significant additional amount of CO2 from the
atmosphere. One possible approach for
slowing the increase in greenhouse gases
concentrations in the atmosphere is to manage
terrestrial ecosystems to conserve or sequester
additional carbon.

Forest ecosystems represent an important
opportunity for conserving and sequestering
carbon because of their large accumulations of
woody biomass. Creating new forests or
restoring degraded  ones can significantly
increase carbon sequestration. Similarly,
agricultural systems, which  cover vast
acreages, can be managed on a yearly basis
to augment the large store of carbon in their
soils.

The Office of Policy, Planning, and Evaluation
at the U.S. Environmental Protection Agency, in
collaboration with the U.S. Forest Service, is
conducting an assessment of land use
management policies that can contribute to
stabilizing U.S. greenhouse gas emissions.
This assessment analyzes potential
greenhouse gas, economic, and other impacts
of land use management policies. The
assessment uses a number of different sectoral
models, which have been, where feasible,
linked to provide a comprehensive evaluation
of the forest and agricultural sectors. Scenarios
that represent the implementation of a number
of policies are being analyzed using the
assembled models. While much  of this work is
still ongoing, this study presents preliminary
results of the following scenario analyses:
    Tree Planting on Marginal Crop and
    Pasture Land
    Conservation Reserve and Wetlands
    Reserve Programs
    Increased Use of Recycled Paper
    Reduced Harvest on National Forest Land
    Increased Use of Biomass Energy
    Modified Agricultural Tillage Practices
    Increased Use of Winter Cover Crops
TREE PLANTING ON MARGINAL CROP AND PASTURE
LAND

Forestation policies that plant trees on
marginal crop and pasture land increase the
acreage devoted to forests in the U.S. These
policies also increase the amount of carbon
stored in U.S. forests.

•  Carbon Impacts. A large-scale tree
    planting program of 12 million acres,
    costing $220 million annually for 10 years,
    would sequester an additional 6.8 million
    metric tons of carbon annually in 2000, and
    16.7 million metric tons in 2010 (see Table
    ES-1). This accumulation would increase
    the size of the U.S. forest carbon pool by
    between almost 500 and 800 million metric
    tons of carbon by 2040, relative to the
    baseline, with approximately one-fourth of
    the total achieved by 2010.

•  Economic Impacts. Since planting trees
    increases the wood available for harvest
    when the timber reaches maturity, future
    prices for timber are lower than in the
                                             ES-1

-------
    absence of the tree planting program.
    Stumpage prices fall most in the South,
    creating benefits for consumers and losses
    for owners of timber. Lower prices for U.S.
    timber result in a significant decrease in
    imports of Canadian lumber by the year
    2040.

•   Other Impacts. Tree planting on marginal
    lands can reduce soil erosion, protect
    watersheds, and conserve biodiversity.

CONSERVATION RESERVE AND WETLANDS RESERVE
PROGRAMS

One policy option using existing programs to
sequester carbon is to extend current and
planned contracts under the Conservation
Reserve Program  (CRP), thereby preventing
the loss of forested land that would otherwise
revert to cropland. Other options would expand
both the CRP and the Wetlands Reserve
Program (WRP) with tree-planted areas.

•   Carbon Impacts:

    • Enrolling 40 million acres under the CRP
    and maintaining this acreage would result
    in an annual carbon accumulation of 11.6
    million metric tons in 2000, and 7.9 million
    metric tons in 2010. This would increase
    the size of the U.S forest carbon pool by
    almost 200 million metric tons in 2015 and
    nearly 350 million metric tons by 2035,
    compared to the likely reversion of CRP
    lands after 1995.

    • Expanding the CRP to 50 million acres
    by planting an additional 10 million acres
    with trees would result in an annual carbon
    accumulation of 31.5 million metric tons in
    2000, and 15.9 million metric tons in 2010
    (see Table ES-1), and would more than
    double the incremental carbon stored
    relative to the base  case. The total carbon
    pool on these lands would increase by
    over 1,100 million metric tons by 2035.

    • Expanding the WRP to five million acres
    would result in an annual carbon
    accumulation of 2.2 million metric tons in
    2000, and 4.4 million metric tons in 2010.
    The forest carbon pool would increase by
    almost 200 million metric tons by 2035.
•   Economic Impacts. Net economic returns
    increase for crop producers, but decrease
    significantly for livestock producers. The
    net result of changes in income and in
    government payments is a decline in
    combined agriculture and livestock industry
    returns of about  $350 million to $450
    million per year.  Under  the CRP scenarios,
    consumers are generally worse off with
    slightly lower consumption and higher retail
    prices, resulting  in about a 2 percent
    increase in total  consumer expenditures.

INCREASED USE OF RECYCLED PAPER

By decreasing the demand for harvested
wood, increased paper recycling has the
potential to increase standing biomass and,
thus, enlarge the forest carbon pool. In this
scenario, the utilization rate for recycled
wastepaper increases to 45% by 2000 and
remains constant through 2040.

•   Carbon Impacts. Increased use of
    recycled paper would result in an annual
    carbon accumulation of 13.5 million metric
    tons in 2000, and 15.6  million metric tons
    in 2010 (see Table ES-1). The forest
    carbon pool would increase by an
    additional 400 to 600 million metric tons by
    2040, an amount comparable to that of
    large scale tree planting programs.

•   Economic Impacts. Because additional
    recycling reduces the demand for
    pulpwood, stumpage prices and revenues
    for producers decline over time under this
    scenario. Declining revenues occur
    predominantly in the softwood markets in
    the South, which produce the  bulk of the
    paper pulp for markets that switch to
    recycled fiber. Lower U.S. timber prices
    reduce softwood lumber imports from
    Canada, which drop roughly to 36 percent
    below base case levels by 2040.

REDUCED HARVEST ON NATIONAL FOREST LAND

Harvest  reductions in this scenario result from
eliminating harvest of old growth volumes in
the Pacific Northwest, protecting spotted owl
habitat in Washington, Oregon and California,
protecting the red cockaded woodpecker in
the South, eliminating below cost timber sales,
                                            ES-2

-------
and eliminating harvesting in existing roadless
areas.

•  Carbon Impacts. Decreases in National
   Forest harvest would be mostly offset by
   increased harvests by other landowners, or
   in other regions. Overall there is little
   change in the total timber inventory and,
   hence, in the total carbon stored on U.S.
   timberland. Offsetting harvests in other
   countries could, however, increase carbon
   storage in  U.S. forests over base case
   levels. Carbon losses in those other
   countries might be greater than the carbon
   savings realized in the U.S.

•  Economic Impacts. Because timber
   inventories are reduced, softwood lumber
   prices increase, prompting  a slight
   decrease in consumption of 16% and an
   increase in Canadian imports of 10% by
   2040.  Higher prices also increase the use
   of more energy-intensive, non-wood
   substitute products.  Analysis of
   international trade linkages suggests that
   higher prices increase harvests in countries
   other than the U.S.  This result could lead
   to economic surplus gains of $1.4 billion to
   timber producers worldwide and a net
   welfare loss of  almost $100 million for U.S.
   producers.

INCREASED USE OF BIOMASS ENERGY

Under this scenario, energy supplied by
fuelwood from existing forests increases from
3.1 Quads in 2000 to 5.4 Quads in 2030, to
meet levels of  bioenergy supply contained in
the National Energy Strategy (NES).

•  Carbon Impacts. Increasing the production
   of energy from biomass has three distinct
   effects on  CO2 emissions: (1) CO2 is
   released when the wood is combusted to
   produce energy; (2) carbon is  sequestered
   during the growth or regrowth  of forests
   harvested  for biofuels; and (3)  carbon
   emissions  are "avoided" when fossil fuel
    combustion is displaced by bioenergy.
    Removing  wood for biofuels initially has the
   effect of reducing the carbon inventory on
   the forest base. Over time,  however,
    regrowth on harvested lands replaces the
    biomass carbon. Increased use of biomass
    energy results  in an annual forest carbon
   accumulation of -7.3 million metric tons in
   2000, and -17.8 million metric tons in 2010
   (see Table ES-1). The real benefit to the
   atmosphere occurs as wood displaces
   fossil fuels. Avoided fossil fuel emissions
   are 2.6 million metric tons in 2000, and 7.5
   million metric tons in 2010. Combined net
   carbon impacts equal -4.7 million metric
   tons in 2000, and -10.3 million metric tons
   in 2010.

COMBINED FOREST POLICY SCENARIOS

Because policies can have offsetting economic
impacts, there are benefits to jointly
implementing forest sector policies.  Combining
scenarios presents an opportunity to evaluate
the carbon, timber, and economic
consequences of a coordinated forest sector
strategy to sequester carbon. The combined
scenarios are:

•  Reduced Harvest and Increased
   Wastepaper Recycling (Combination 1)
•  Combination 1 with Increased Production
   of Energy from Biomass (Combination 2)
•  Combination 2 with Large-scale Tree
   Planting (Combination 3)

The combined scenarios are projected to have
the following effects.

•  Carbon  Impacts. The carbon impact of the
   combination 1 scenario is similar to that of
   the increased paper recycling scenario;
    10.6 million metric tons of carbon are
   accumulated annually in 2000 and 13.0
   million metric tons are accumulated
   annually in 2010.  In addition:

    • Adding the biomass energy scenario
    (combination 2) reduces carbon
   significantly, even when avoided fossil fuel
    emissions are taken into account. Net
    annual carbon accumulation equals 6.8
    million metric tons in 2000, and 3.7 million
    metric tons in 2010.

    • Adding large-scale tree planting to this
    scenario (combination 3) illustrates the
    benefits of combining a bioenergy strategy
   with plantations established to grow biofuel
    stocks. Net annual carbon accumulation for
    combination 3 rises to 12.7 million metric
                                             ES-3

-------
    tons in 2000 and 18.5 million metric tons in
    2010.

•   Economic Impacts. Combining increased
    recycling with reduced harvests on public
    lands creates offsetting impacts on timber
    prices.  Stumpage prices for combination 1
    are higher than for the recycled fiber
    scenario and lower than the harvest
    reduction scenario. Because of the strong
    influence of additional recycling, prices are
    lower than in the base case.  In addition:

    • Adding a large  biomass energy program
    that draws on existing resources in the
    forest sector increases prices over the first
    scenario, creating economic gains for
    stumpage suppliers in the combination 2
    scenario.

    • Introducing large-scale tree planting
    reduces prices, relative to combination 2.
    Because combination 3 includes the
    reduced harvest scenario and biomass
    energy scenario,  however, which draw
    down the timber inventory, prices are
    above those for the tree planting scenario
    alone.

    Thus, the presence of multiple activities in
    the combined scenarios reduces the
    adverse economic impacts of individual
    scenarios.

MODIFIED AGRICULTURAL TILLAGE PRACTICES

It is estimated that agricultural soils store 1.5
trillion metric tons of carbon, or roughly twice
the amount of carbon in the atmosphere.
Under this scenario, tillage practices on highly
erodible lands were changed to reduced-till
and no-till practices in order to meet soil
conservation targets.

•   Carbon Impacts. Changing agricultural
    practices, that continue the trend towards
    conservation tillage, would result in
    average annual soil organic carbon
    accumulations of up to 2.5 million metric
    tons above base  case levels (see table
    ES-1). Estimated  soil carbon accumulation
    rates are very sensitive to projected crop
    yields. Soil carbon impacts on individual
    areas of the  U.S.  can differ significantly
    from national averages due to variation in
    dominant crop rotations, other production
    practices, climate, and soil conditions.

•   Economic Impacts. Economic impacts
    depend primarily on assumptions about
    short term crop yield decreases due to
    unfamiliarity with the new tillage practices.
    Without yield adjustments, net returns
    (gross revenues less variable costs) under
    the low conservation scenario increase by
    $0.67 per acre, or 0.8 percent. Net returns
    under the more aggressive high
    conservation scenario increase by $4.06
    per acre, or 5.4 percent. With yield
    adjustments for implementing conservation
    tillage practices, net returns decrease by
    $3.02 per acre or (3.4 percent) for the low
    conservation scenario, and decrease by
    $2.93 per acre or (3.3 percent) for the high
    conservation scenario.

•   Other Impacts. As an additional  benefit,
    soil erosion decreases by 20 to 30 percent
    with increased use conservation tillage
    practices.

INCREASED USE OF WINTER COVER CROPS

This scenario introduces winter cover crops to
areas with favorable conditions (areas with
proper climate, crop rotations, and length of
growing season), thereby increasing total
biomass  production and associated soil
carbon accumulation on agricultural lands over
the course of the year.

•   Carbon Impacts. Although lands targeted
    for winter cover crops account for only 5 to
    10 percent of the total land area in the
    study region, there are significant
    increases in the level of soil organic
    carbon. Increased planting of winter cover
    crops results in an average annual soil
    organic carbon accumulations of up to 3.5
    million metric tons above base case levels
    (see Table ES-1).

•   Economic Impacts. Average net economic
    returns to farms decreased by slightly over
    1  percent as a result of these cover crop
    policies.

•   Other Impacts. Some cover crops, such as
    hairy vetch, have the additional benefit of
    fixing nitrogen in the soil.
                                             ES-4

-------
                              Table ES-1
Annual Carbon Accumulation for the Forest and Agriculture Sector Scenarios,
                  Decadal Intervals (million metric tons)

Tree Planting on Marginal Crop and
12 million acres,
optimal distribution
6 million acres,
optimal distribution
4 million acres,
historic distribution
Conservation and Wetlands Reserve
40 million acres in CRP
50 million acres in CRP
5 million acres of Wetlands Forest
Increased Use of Recycled Paper
Increased Energy from Biomass
Forest Carbon
Avoided Fossil
Fuel Emissions
Net Carbon Flux
Combined Forest Policy Scenarios
Combination 1:
NF harvest reduction and recycling
Combination 2:
Combination 1 and biomass energy
Forest Carbon
Avoided Fossil
Fuel Emissions
Net Carbon Flux
Combination 3:
Combination 2 and tree planting
Forest Carbon
Avoided Fossil
Fuel Emissions
Net Carbon Flux
Modified Tillage Practices
Winter Cover Crops
2000
Pasture Land
6.8
4.1
3.5
Programs
11.6
31.5
2.2
13.5

-7.3
2.6
-4.7

10.6

4.2
2.6
6.8

10.1
2.6
12.7
2.5
3.5
2010

16.7
9.6
6.5

7.9
15.9
4.4
15.6

-77.8
7.5
-10.3

13.0

-3.8
7.5
3.7

11.0
7.5
18.5
2.5
3.5
2020

19.4
10.8
7.5

7.7
34.5
7.0
10.0

-29.6
73.2
-16.4

8.8

-23.8
73.2
-10.6

-6.6
73.2
6.6
2.5
3.5
2030

13.0
6.2
5.5

7.2
31.6
4.6
4.4

-27.7
27.5
6.4

3.7

-24.2
27.5
3.3

-72.7
27.5
14.8
2.5
3.5
                                  ES-5

-------

-------
1.  INTRODUCTION
    Rising atmospheric levels of greenhouse gases, which are the result of anthropogenic sources of
emissions, are likely to induce changes in the earth's climate over the coming decades. Greenhouse
gas-induced climate change, which is projected to result in rising global averagejemperatures and
rising sea levels, may have associated impacts on energy demand, water resource quality, coastal
property and ecosystems, and the commercial forestry and agriculture sectors. There remains
considerable uncertainty about the rate and magnitude of possible climate change, as well as about
the physical impacts associated with  such change. There is, however, an emerging consensus that
policies to stabilize or reduce emissions of greenhouse gases — which include carbon dioxide (CO2)
and other radiatively important trace gases, such as methane (CH^,  nitrous oxide (N2O),
hydrofluorocarbons (HFCs), and perfluorocarbons (PFCs) — should be explored (see Box 1-1).
    The vegetation and soil in terrestrial
ecosystems contain about three times as much
carbon as is in the atmosphere, and play a
major role in the movement of CO2 and other
greenhouse gases into and out of the
atmosphere. For example, carbon may be
released to the atmosphere as CO2 or CH4
when trees or other vegetation are harvested or
decay, or when soils are tilled. Thus,  increasing
the amount of standing biomass, or changing
cultivation practices, can increase the size of the
carbon sink and decrease concentrations of
greenhouse gases in the atmosphere. One
proposed approach, therefore, for slowing the
increase in atmospheric CO2 is to manage
terrestrial ecosystems to conserve or sequester
(remove and store) carbon.

    Forest ecosystems present particularly
significant opportunities for carbon conservation
and sequestration because of their large accu-
mulations of woody biomass. In 1990, U.S.
forests recaptured approximately 9 percent of
the CO2 emitted in the U.S. from the  combustion
of fossil fuels (USEPA 1994). Strategies in the
forest and agriculture sectors have the technical
potential to recapture a significantly greater
portion of U.S. CO2 emissions in the future.

    The Office of Policy, Planning, and
Evaluation at the U.S. Environmental Protection
Agency, in collaboration with the U.S. Forest
Service, is conducting an assessment of land
use management policies that can contribute to
stabilizing U.S. greenhouse gas emissions. A
broad set of policies in the agriculture and forest
sectors that could potentially increase the
capacity of terrestrial carbon storage include:
                 £ox 1-1
      Actions to Mitigate Climate Change

    tn 1992, the United Nations conference on
Environment and Development (UNCED) formu-
lated the Framework Convention on Climate
Change, which has since been signed by the U.S.
and at least 160 other countries. On Earth Day
1993, President Clinton announced that the U,&
woufd participate Tn climate change mitigation
efforts and, in October 1993, released the U,&
Climate Change Action Plan (CGAP), which has as
Its goaf returning U,S.  emissions of greenhouse
gases to 1990 levels by the year 2000.

    Emissions of greenhouse gases are generally
measured in million metric tons of carbon
equivalent (MMTCE)* MMTCE provides a
consistent measure to compare emissions of
different greenhouse gases that accounts for the
relative global warming input of each gas. In 1990,
U,S, net greenhouse gas emissions equalled 1,444
MMTCE, Of this total,  1,233 MMTGE (85 percent)
were net carbon released as C02 emissions, 162
MMTGE (11 percent) were methane (CHA 30
MMTCE (2 percent) were N2O, and 19  MMTCE
(1 percent) were HFCs or PFCs,

    tn the absence of the CCAP, the
Administration projects that net greenhouse gas
emissions would grow to 1,568 MMTCE by 2090,
Under the CCAP, which Includes measures to
reduce various sources of emissions as well as
increase carbon sequestration by forests, the
Administration estimates net greenhouse gas
emissions of 1,459 MMTCE* This estimate reflects
reductions by the year 2000 of 109 MMTCE, or 7
percent of estimated emissions tn the year 20DO,

Sources: USEPA 1994; USDOS 1994
    enlarging the acreage of trees by
    encouraging rural and urban tree planting and adopting long term fire management strategies;

-------
•   increasing the rate of tree growth on existing forest land by improving timber management
    practices or using improved seedling stock;

•   using biomass-based fuels, which reduce carbon emissions from fossil fuels, and do not
    themselves have a net impact on long-term atmospheric carbon;

•   increasing utilization rates for recycled paper, thereby reducing demand for pulpwood and
    maintaining and increasing timber inventories;

•   conserving standing primary and old-growth forests as stocks of biomass by restricting  harvesting
    or setting aside land;

•   reducing soil disturbances by  modifying tillage practices and thereby increasing soil organic
    carbon; and

•   planting winter cover crops to increase biomass on agricultural lands.

    This document summarizes the results of a number of studies conducted over the last three years
to examine forestry and  agricultural policies, including many of those described above, and their
implications for carbon storage and emissions of greenhouse gases (see Box 1-2). Understanding not
only the greenhouse gas implications but also the socioeconomic effects of these policies requires
answering a series of key questions:

•   To what extent can the policy  or program sequester or offset greenhouse gas emissions?

•   How much will the policy or program cost the government?

•   What are the economic and social impacts on  the forest and agriculture sectors nationally and
    regionally?

•   How effective are the policies or programs when combined?

•   What are the impacts of these domestic policies on global trade, and of global trade on these
    policies?

    The results of the studies reported in this document build on a body of literature, both scientific
and policy-analytic.  Past analyses of the potential impacts of domestic forest and agriculture
management practices have concentrated on one  or two of a multitude of factors — an individual
policy, effects  on soil carbon, or program cost. This study evaluates the full range of economic
implications and quantifies the likely physical effects for a variety of potential strategies. Together with
the past literature, the studies reported  here represent an initial step towards building the volume of
information necessary to comprehensively assess viable policy options in the forest and agriculture
sectors for reducing/offsetting greenhouse gas emissions.

    This study has assembled a set of models which,  collectively, evaluate the impacts of various
forest and agriculture sector strategies designed to reduce or sequester carbon emissions.  Following
a brief description (in Section 2) of carbon processes, land  use change, and the scenarios examined,
Section 3 presents the models used in the study. Sections 4,  5, and 6 present the results of the
scenario analyses. Section 7 concludes with a brief discussion of research directions.

-------
                                            Box 1*2
                            Studies Conducted for this Assessment

  The Forest Sector Carbon Budget of the United States:  Carbon Pools and Flux Under Alternative
  Policy Option*. May 199& Edited by D>P. Tomer, J*J, Lee, G.4> Koerper, and J>R Barker, US. EPA
  Environmental Research Laboratory at CorvalliS, Oregon. EPA report EPA/600/3-93/093. Supplemented by
  additional scenario analyses.

  Alternative Simulations M, Perest-Garcia, Center for
  International Trade in Forest Products at the University of Washington."

  Economic and Resource Impacts of PoUcles to Increase Organic Carbon tn Agricultural Soils. May
  1993, Report  prepared by A Bouzaher, OJ, Holtfcamp, ft. Reese, and J. Shogren, Center for Agricultural and
  Rural Development at lowa Slate University,

  Long Term Economic Consequences of Alternative Carbon  Reducing Conservation and Wetlands
  Reserve Programs: A BLS Analysis. May 1993. Report prepared by R. Reese, A. Bouzaher, and J.
  Shogren, Center for Agricultural and Rural Development af Iowa State; University,

  Assessment  oitAjterjtatlva Management Practices and Policies Affecting Soft Carbon In
  Agroecosystems  of the Central United States. April 1994. Report prepared by AS. Oonigian, A.S.
  Patwardhan,  and R. Chinnaswamy, AQUA TERRA Consultants; T.O, Bacnwelt, Jr. and R.B, Jackson, IV, O:S.
  EPA Environmental Research. Laboraiory at Athens, Georgia; K.B, Wejnrich and A,L, Rowett, computer
  Sciences Corporation; and C.V.  Cole, Natural Resources Ecology Laboratory. EPA/60G/R-94/067.

  Planting Trees on Agricultural Lands; The Costs and Economic Impacts of GRP Reversion. December
 ' 14, 1994, Report prepared by JM Catlaway, RCG/Hagler Batlty and 6.F, McCar), Texas A&M University,

  Carbon Sequestrat/on and N20 Emissions from Soils: A Model Simulation Study for Seven Agricultural
  Sites In the Central V.S, Ustoff the DNDC model. October 7,1993> Draft report prepared by C< Li and A.
  Cialella, The Bruce Company.
2.  BACKGROUND: THE CARBON CYCLE, HISTORIC LAND USE CHANGE, AND
    MITIGATION STRATEGIES

    Rising levels of atmospheric CO2 may increase global temperatures through the "greenhouse"
effect, which in turn may raise sea level and have serious domestic and international implications for
human health, ecosystems, productive resources, and coastal areas. The atmospheric concentration
of CO2 has risen from approximately 275 parts-per-million (ppm) in pre-industrial times to 350 ppm
today, and is currently increasing at a rate of approximately 1.5 ppm per year. Projections suggest
that the global average temperature will warm by 1.5 °C to 4.5 °C (2.7 °F to 8.1 °F) by the first half of
the next century (IPCC 1992).

-------
    The potential consequences of increasing atmospheric carbon levels are serious enough that
 governments are considering programs to stabilize or reduce anthropogenic emissions. Increasing
 forest carbon sequestration can play an important role in these efforts. To develop effective policy
 options, an understanding of the global carbon cycle, and how anthropogenic carbon emissions may
 affect it, is necessary. This section presents background information on the role of terrestrial
 ecosystems in the global carbon cycle, describes how historic land-use changes have determined the
 current status of U.S. forests, and presents the policy scenarios that are evaluated in this report.

    2.1  THE GLOBAL CARBON CYCLE AND CARBON SECTOR BUDGETS

    The global carbon  cycle is the movement of carbon among the  atmosphere, terrestrial biosphere,
 and ocean. Current estimates suggest that the oceans store by far the largest amount of carbon —
 roughly 40,000 billion metric tons of carbon (see  Figure 2-1). Soil is the next largest  reservoir, or pool,
 followed by the vegetation and the atmospheric carbon reservoirs, which are about equal in
 magnitude. Although the terrestrial biosphere (vegetation and soils) stores much less carbon than do
 the oceans, the  amount of carbon that cycles annually between the terrestrial biosphere and the
 atmosphere is similar to that which cycles between the oceans and the atmosphere. Together, the
 combined annual flux between these two pools and the atmosphere is about 20 percent of the total
 atmospheric carbon pool.
                                   Figure 2-1.  The Global Carbon Cycle
                                            Atmosphere
                                              750 Pg
                                            *• 3 Pg per year
                                                             92 Pg
                                                            Biological
                                                             and
                                                        Chemical Processes
                            60 Pg
                          Respiration,
                         Decomposition
                                                                       Chemical Processes
          Sources: IPCC 1994; IPCC 1992; Schneider 1989
    From the perspective of possible global climate change, the concern is that more carbon, primarily
in the form of CO2, is accumulating in the atmosphere than is being removed through plant
photosynthesis and other biological processes. It is estimated that the atmosphere is gaining about 3
billion metric tons of carbon annually. The two principal sources of atmospheric CO2 increases are
fossil fuel combustion and deforestation, which contribute 5.5 billion and 1.1  billion metric tons of
carbon, respectively, worldwide. Further, trends suggest that rates of anthropogenic emissions of CO2
from these two sources are rising. Terrestrial vegetation is an important component of the carbon

-------
cycle, and policies designed to enhance its capacity as a carbon pool, or reduce its emissions of
greenhouse gases, can potentially have a significant impact on atmospheric levels of CO2.

    FOREST SECTOR CARBON BUDGET*

    A carbon budget is a bookkeeping system for tracking the amount of carbon in various reservoirs
("pools") and the amount of carbon transferred among the reservoirs and the atmosphere ("fluxes").
The forest sector carbon budget has three major pools: the organic matter of forest ecosystems, the
products derived from forests, and products that are landfilled, incinerated, or otherwise discarded
(see Figure 2-2). The main uptake of carbon into forests is through photosynthesis, the fixation of
atmospheric CO2 by green plants. Carbon loss from forests occurs primarily from burning and from
the harvest of wood by humans. The net flux of carbon from the atmosphere to the forest is the
difference between carbon uptake from photosynthesis and carbon release from respiration by all
forest organisms, including decomposers. The "net accumulation" of carbon by forest ecosystems is
the net flux minus carbon removed by harvest and burning. Detailed forest sector inventories
developed periodically by the U.S. Forest Service provide an accurate basis for estimating forest
carbon storage. The analyses conducted for this report model the impacts of various policies  on forest
inventories over time and estimate the associated changes in carbon  storage.
                       Figure 2-2.  The Forest Sector Carbon Budget
CO2

CO2

{Photosynthesis * Burning,
1 Decomposition
Forest
Ecosystem
—
Harvest
—
CO2

# Incineration
Products

{Respiration. I
Burning V
COj

Landfills

                                                                       Decomposition
             Source: Turner, etal. 1993
    Unlike the forests, there are no systematic inventories of the product and disposal pools, or of the
net annual transfer of material into or out of these pools. Some products, such as furniture and lumber
for housing, are a relatively long-lived store for carbon. Once these and other products, such as
newspaper and paper packaging, enter disposal pools, much of the material decays and the carbon is
emitted to the atmosphere in the form of greenhouse gases. The length of time over which material
decays depends on the type of material and on factors specific to the landfill. In some cases, a
portion of the carbon may effectively be permanently stored in the landfill. Transfers of harvested
   1
     This section is based primarily on Turner, et al. 1993.

                                               5

-------
forest biomass into product and disposal pools can, therefore, have a significant impact on the timing
and magnitude of carbon accumulation in the entire forest sector.

    AGRICULTURE

    Vegetation and soil in the agricultural sector also store carbon. Soil organic matter, in the form of
live crops in the ground, crop residues, and other dead and decaying plant material and soil micro-
fauna, account for the carbon found in agricultural soils. Soil carbon accumulation is determined by
the balance of biomass inputs from primary production (e.g., crop residue) and carbon loss from
decomposition and erosion (Donigian, et al. 1994). Thus, agricultural management practices such as
crop selection, crop residue treatment (e.g., burning, removal), and tillage, have a large impact on soil
carbon accumulation. For example, tilling soil during the planting season disturbs some of the soil
organic matter and promotes decomposition, releasing CO2 into the atmosphere. Management
practices that increase production, such as introducing winter cover crops into an annual crop
rotation, can also  increase soil carbon accumulation.

    2.2 HISTORIC LAND USE CHANGE: U.S. FORESTS*
    The history of the conversion of original
forest to its current state is important for a
complete understanding of trends and modelling
projections of carbon flux and storage. In 1630,
the continental U.S. had about 822 million acres
of forest land, characterized by a high
proportion  of very large trees and in
approximate equilibrium with respect to growth
and mortality. With  the exception of pockets in
the western U.S., most of the pre-settlement
forest has now been cleared and developed, or
replaced with agricultural land  or successional
forest. Productive forest land (see Box 2-1) is
estimated at 537 million acres,  with another 200
million acres  of urban and marginal forest land
(now called other forest) (Haynes, et  al. 1994). A
division of the continental U.S.  into three regions
— North, South, and West — broadly  captures
differences in the use and management of the
pre-settlement forest lands. Current and historic
carbon storage on these lands reflects the
different land use patterns in each region (see
Figure 2-3) .3

    NORTHERN REGION

    By the  end of the 19th century, virtually all
northern forests had been cleared for
agricultural use or heavily logged for timber
products, with the exception of extreme northern
                  Box 2-1
   Definition of Timberland and Woodland

Forest land in the US. Is differentiated between
timberland> reserved timbertancif and other forest.
Timberland is forest (and that is producing at is
capable of producing crops of Industrial wood and
that is not withdrawn from timber utilization by
statute or administrative regulation. (Mote; areas
Qualifying as tlmberland are capable of producing
in excess of 20 cubic feet per acre per year of
Industrial wood in natural stands. Currently
inaccessible and inoperable areas are included.}

Reserved timberiand refers to land having the
productive capacity of timberiand but which Is
withdrawn from timber utilization by statute or
administrative regulation,

Other forest land f woodland") includes fofeSt land
other than timberiand ana reserved timberiand* if
includes available and reserved unproductive forest
land, which Is incapable of producing 20 cubic feet
per acre per year of industrial wood under natural
conditions because of adverse site conditions such
as sterile soils> dry  climate, poor drainage, high
elevation, steepness, OF cockiness.

Source: Turner, etaL 1993
     This section is based primarily on Birdsey and Heath 1993.

   3 In terms of the regions identified subsequently in Figure 4-2, the West is composed of the Pacific Northwest
and Southwest and the Rocky Mountains; the North combines the Northeast and the North Central; and the South
contains the Southeast and South Central.

-------
                          Figure 2-3. Carbon Storage on U.S. Timberland, by Region
40 -
35 -

J 30 -
o
° 25 -
g
in
g 20 -
3| 15 -
c
_o
S 10 -
5 -
0
195


f *• *
f -

~
-


O 1960 1970 1980
f Total U.S. _--.- 	 ""





-North
r South
I
West
t 1 1 1 	 r
1990 2000 2010 2020 2030 2040
Year
               	Historic-
                                                                -Projected-
                Source: Birdsey and Heath 1993
areas of Maine and the Great Lakes States and some inaccessible areas in the Appalachian
mountains. Beginning in the mid-nineteenth century and accelerating in the 20th century, marginal
agricultural land reverted to forest and produced a dense stocking of trees, with much of the area now
covered by trees in the 25 to 65 year age classes. These forest lands of mixed species are in the
middle of a period of rapid growth that can be sustained for several  more decades before reaching a
period of declining growth. In the near future, therefore, forest biomass is  expected to continue to
increase.

    SOUTHERN REGION

    As in the North, only scattered fragments of pre-settlement forest remain. The current forest has
been regenerated either naturally or artificially on  marginal cropland, pasture, or cutover forest land,
and many areas are being cut for the third time and regenerated to a fourth timber stand. The
intensively managed southern forests have the youngest age class distribution in the U.S., with a large
proportion in the 25 to 45 year age classes. Because of intensive management and a continuing shift
of industrial timber supply to the South, most of these forests are not expected to reach biological
maturity before further harvest and regeneration. Forest biomass is expected to remain constant or
decline slightly over time.

    WESTERN REGION

    Although the West has not experienced the major land use shifts characteristic of other regions,
forest disturbance has dominated the landscape as the original forests have been logged and
converted to second-growth forests. The remaining old-growth in the West comprises less than 2

-------
percent of the pre-settlement forest. Both private and public western forests have a relatively flat age-
class distribution, but with a large component of older forest classes. It is expected that western public
forest lands will be increasingly reserved from timber cutting, producing a long-term net gain in forest
biomass on public lands. As harvesting shifts from public to private lands, forest biomass on private
lands will correspondingly decline.

    2.3 HISTORIC LAND USE CHANGE:  AGRICULTURE

    The extensive conversion of land to agriculture, which began just after the turn of the century,
resulted in substantial decreases in soil organic carbon relative to native ecosystems. Traditional
agricultural practices involved increased disturbance of soil, reduced plant cover during fallow  periods,
and the removal or burning of crop residues, which minimized carbon inputs and promoted organic
matter decomposition (Donigian, et al. 1994). These practices led to an estimated 47%  decrease in
soil organic carbon between 1907 and 1950. This release represents the second largest
anthropogenic source contributing to historical increases in CO2, after fossil fuel combustion (Post, et
al. 1990).

    Following this period of decline, soil organic carbon remained fairly constant for the next two
decades as modern crop management practices were developed. These management  practices have
decreased the level and intensity of tillage, and have increased the amount of crop residues that are
returned to the soil. Soil carbon models that reflect changes in agricultural practices indicate that soil
organic carbon levels have been steadily increasing since 1970, although this  trend has yet to  be
confirmed at a national or regional level (Donigian, et al. 1994).

    2.4 OVERVIEW OF SCENARIOS

    There are a number of trends in the forest and agriculture sectors, as well as potential new
programs, that could increase the next accumulation of carbon in forest and agricultural ecosystems.
The scenarios considered in this report, which are summarized in Box 2-2, represent  a  range of
plausible trends, new programs, or expanded existing programs. Three of these scenarios (large-scale
tree planting, and tree planting on conservation or wetlands reserve program lands) increase the
amount of carbon stored on U.S. forest land by increasing the acreage of standing trees.4 Two
additional scenarios expand the forest carbon pool by reducing  harvesting on existing stands
(reduced National Forest harvest and increased wastepaper recycling). A sixth scenario increases the
use of wood for biofuels, and additional scenarios combine policies.

    In the agriculture sector, modifying existing agricultural practices can increase soil organic  carbon
levels. Employing tillage practices that do not disturb the soil as much as conventional  tillage, and
planting winter cover crops, can in some cases increase carbon sequestration in agricultural soils, and
provide other benefits such as reduced erosion or increased nitrogen fixation.


3.  METHODOLOGY

    This report analyzes potential economic, greenhouse gas, and other  impacts of land use
management policies. The assessment uses a number of different sectoral models, which have been|
where feasible, linked to provide a comprehensive evaluation of the forest and agricultural  sectors (see
Box 3-1). The relationships among these models  are illustrated in Figures 3-1 and 3-2.
     The Conservation Reserve Program provides incentives to remove highly erodible land from agricultural
production by paying for conversion costs, and providing rental payments over 10 year contracts. The WRP
provides incentives to restore and protect wetlands that have been fully or partially converted to agricultural
production, by paying for conversion costs and compensating owners for the land value.

                                               8

-------
                                 ,  ,  -;    BOX2-2
                                    Summary of Scenarios

  Increasing Forest Carbon Storage

      Afforestation >— Tree planting on marginal crop and pasture land Two alternative funding levefs for tree
      planting programs are analyzed ($11D mlion and $220 rnlion annually lor to years, or about 6 to 12
      million acres).

      Conservation Reserve Program (CRP) — Branding and maintaining the current CRP plans, thereby
      Increasing fbrestland acreage above the current program. Two alternative proposals are analyzed {a 40
      million acre CRP and a 50 million acre CRP}<

      Wetland Reserve Program (WRP) — introducing a 5 million acre reserve of wetlands targeted to re-
      establish bottomland hardwood forest stands on suitable agricultural lands,

      Increased Recycling — The utilization rate for recycled paper increases to 45 percent of total fiber
      production In tie year 2000 and remains constant through 204a

      Reduced National forest Harvest — Harvest on National Forests declines relative to the base case to
      reflect additional reserved timberiand and set-astdes,

      Wood Energy fj>/0f«efej — Fuelwood consumption for bjofueis increases as described in the National
      Energy Strategy.

      Combination Scenario 1 — Combines reduced National Forest harvest with recycling.

      Combination Scenario 2 — Combines reduced National Forest harvest, recycling,, and biofuels.

      Combination Scenario 3 — Combines reduced National Forest harvest, recycling, biofuels, with $220
      million tn annual funding over 10 years for tree planting programs.
                                           *                             •*
  Strategies in the Agriculture Sector

      Modified Tillage Practices — Increased use of Reduced'Till and No-Till practices on agricultural land,

      Winter Cover Crofts — Introduce winter cover crops where climate, crop rotations, and length  of growing
      season allow.
    For the forest sector, changes in forest inventory on private timberlands are estimated by linking
an economic model (TAMM), which estimates tree harvest volume by region, to a forest inventory
model (ATLAS), which tracks forest age class distribution by forest type and productivity class. TAMM
ensures that the economic consequences of implementing the various policy options are allowed to
affect the carbon budget. The resulting forest inventory at any point in time is run through two
alternative carbon models (FCM and FORCARB), to estimate total forest carbon. The model
HARVCARB is used to estimate changes in the amount of carbon in forest products and in disposal
pools from material harvested from 1990 to 2040.

    The basic forest sector analyses using TAMM/ATLAS and the carbon models are supplemented in
several ways. First, because tree planting on marginal crop and pasture land may affect prices and
production in the agriculture sector, additional results from two models of the U.S. agricultural sector,
ASM and BLS, provide insights into the potential impacts of tree planting policies. Second, because
the U.S.  exports and imports raw timber and forest products, policies to expand the U.S. timber
inventory can affect world markets. The CINTRAFOR global trade model is used to evaluate these

-------
                                            BOX 3-1
                       Summary of Key Models Used for this Assessment

   Forest Sector Models

      Tlmberland Assessment Market Model (TAMM) — economic model of the U,S. forest sector, developed
      by the U.S. Forest Service.

      Aggregate Tlmhertand Atse&ment System (ATLAS) —forest inventory change modal for private
      timberland in the U.S,, developed1 by the U.S, Forest Service and finked to TAMM,

      North American Put ft and Paper (NAPAP} mode/ — sectoral moctel of demand,  supply and
      technology for the pulp and paper sector of U.S, and Canada, developed by the U.S. Forest
      Service.

      Forest Carbon Modet9 (FCM and FOHGARty — models of the carbon contained in forest
      ecosystems, developed &y the U.S, Environmental Protection Agency and the U.S. Forest Service,
      respectively,

      HARVCARB -~ model of post-harvest carbon, tracing the flows of carbon through processing, use,
      and disposal, developed at the institute for Forest Analysis, Planning, and Policy.

      CINTRAFOR Global Trade Model (G&THII) — global forest sector model, developed by the
      International Institute for Applied Systems Analysis (IIASA) and the University of Washington's Center
      for International Trade fn Forest Products (CiNTRAPOR).

   Agriculture Sector  Models

      Basic Unked Systems (BUS) — model linking 34 national and regional models to simulate world
      agricultural production and trade, developed by Iowa State University's Center for Agricultural and
      Rural Development (CARD).

      Resource Adjustment Modeling System (RAMS) — model of regional agricultural production,
      developed by  Iowa State University's Center for Agricultural and Rural Development (CARD),

      Denltrlflcatlon and Decomposition (DNDC) Model — model of carbon and nitrogen dynamics In
      agricultural soils, developed by the Bruce company and the university of New Hampshire,

      CENTURY - modei of the dynamics of soil organic matter in agricultural lands and grasslands,
      developed by  Colorado State University in collaboration with Michigan State University.

      Agriculture Sector Model (ASM) — model of the U.S. agriculture sector, developed at the
      Agricultural Economics Department of Texas A&M University.
impacts. Last, to analyze the impacts of increased waste paper recycling, the NAPAP model provides
detail on the North American pulp and paper sector.

    To evaluate the economic and carbon impacts of change in agricultural practices, the analysis
combines models of agricultural production, agricultural markets, and physical processes (such as the
dynamics of carbon and nitrogen in soils). A detailed regional model (RAMS) of agricultural production
methods is used to evaluate the impacts on production of alternative crops, tillage practices, and
rotations. This regional production model is linked to two models (DNDC and CENTURY) of the
physical processes in agricultural soils to capture soil dynamics, fluctuations in the levels of soil
organic carbon, and changes in nitrous oxide emissions. Finally, the impacts of these policies on
                                               10

-------
                        Figure 3-1. Forest Sector Models
                                Policy Triggers
    CGTM
Global Economic
  Trade Model
     I
   TAMM
 Forest Sector
Economic Model

                                    ATLAS
                                 Forest Inventory
                                     Model
                                      I
                                      FCM
                                   FORCARB
                               Forest Carbon Models
                                                                   Fuelwood
                                                                     Model
                                                                    NAPAP
                                                                  Pulp & Paper
                                                                  Sector Model
                                  HARVCARB
                                  Model of Post-
                                  Harvest Carbon
                     Figure 3-2. Agriculture Sector Models
                                 Policy Triggers
                                      I
                                   Land Use
                                   Changes
       ASM
  Agriculture Sector
  Economic Model
     BLS
Agriculture Sector
 Economic Model
                      FCM
                Forest Carbon Model
    RAMS
Agriculture Sector
 Economic Model
                                                                      1
                                                                   CENTURY
                                                                 Soil Carbon Model
                                                                      DNDC
                                                                CENTURY Validation
                                        11

-------
 agricultural markets are evaluated using a general equilibrium model (BLS) which accounts for the
 linkages between agricultural markets and the non-agricultural markets, regional and international
 feedbacks related to trade, and impacts on different commodities, including the major cash crops and
 livestock.

    3.1  FOREST SECTOR MODELS

    The Timber Assessment Market Model (TAMM) is a model of the U.S. forest sector developed by
 the U.S.  Forest Service, which provides an integrated structure for examining the behavior of regional
 prices, consumption, and production in stumpage and product  markets (Adams and Haynes 1980;
 Haynes and Adams 1985). It is designed to provide long-term projections of price, consumption, and
 production trends and to simulate the effects of alternative forest policies and programs. TAMM
 includes six lumber and plywood demand regions and eight product and stumpage supply regions
 within the contiguous U.S. Projections under TAMM are dependent on exogenous inputs from other
 models: pulp fiber requirements (from the NAPAP model described below), and trade and fuelwood
 projections. The background assumptions for any TAMM model run include projections of future forest
 products markets (based on forecasts of the major determinants of product demand, and of timber
 and product supply). On the demand side, the major determinants are future growth in aggregate
 economic activity, as measured by gross national product (GNP), and in certain  key end-uses for
 forest products, such as new residential construction and the repair or alteration of existing dwellings.
 It is important to note that TAMM does not account for changes in managerial behavior in response to
 price signals and likely  overestimates any adverse economic impacts due to tree planting or reduced
 harvests (Winnett, et al. 1993). TAMM also does not account for forests that are managed for non-
 commodity values.

    The Aggregate TimberLand Assessment System (ATLAS)  projects inventories for all private
 timberland in the U.S. (Mills and Kincaid 1992). Developed by the U.S. Forest Service, ATLAS is an
 age-based, yield table model that projects acres by detailed strata for time intervals consistent with the
 inventory stand-age classes. In the model, the inventory is represented by acreage cells classified by
 region, ownership, management type, management intensity, and age class. Inputs to the model
 include estimates of harvest, acreage shifts, management alternatives, and growth parameters. A
 major attribute of the model is that it can simulate shifts in management intensities and the
 consequent changes in yields. It can also account for both partial harvests and thinning.

    The North American Pulp and Paper (NAPAP) sector model uses linear programming to solve for
 market equilibrium in spatially specific markets. NAPAP represents technological  options as economic
 choices via activity analysis, with production capacity modeled for many competing production
 processes. The pulp and paper sector is modeled as eight demand regions for 14 categories of final
 products and five supply regions, which together provide sufficient detail to  capture market equilibria
for all principal commodities of the pulp and paper sector in  North America (U.S. and Canada). The
 model includes estimates of regional supply functions for manufacturing inputs; regional demand
functions for commodity outputs; technological coefficients describing production possibilities in terms
of inputs per unit of product output in various production processes for each commodity; other
 manufacturing costs by process; transportation costs for various commodities and regions; regional
 production capacities for the various processes; regional constraints on recovery of paper for
recycling; import and export ad valorem taxes; and monetary exchange rates (between U.S. and
Canada).

   The CINTRAFOR Global Trade Model (CGTM) is a global forest sector model developed  by the
International Institute for Applied Systems Analysis (NASA) and the University of Washington's Center
for International Trade in Forest Products (CINTRAFOR). CGTM  provides forecasts of production,
consumption, prices, and trade for eight different forest product commodities, in 43 timber supply
regions and 33 demand regions (Perez-Garcia 1993, Cardellichio et al. 1989). Forecasts of market
behavior are subject to  regional constraints including  resource availability and production capacity.
                                             12

-------
CGTM searches for a spatial (or global) equilibrium, which is the point where the amount supplied by
several log and commodity markets equals the amount demanded by all consumers.

    The Forest Carbon Model (FCM), developed by the U.S. Environmental Protection Agency, is
used to generate estimates of forest carbon from the inventories generated by the TAMM/ATLAS
analyses (Turner, et al. 1993). The yield tables in ATLAS are used as the basis for constructing stand-
level carbon budgets. Carbon pool size per unit area per age class is estimated for living tree
biomass, woody debris, understory vegetation, forest floor, and soil components of the forest.
Separate stand-level carbon budgets are prepared for each of 422 yield tables supplied by the USDA
Forest Service. Carbon storage at the regional and national level is then determined by coupling these
stand-level carbon budgets to forest inventory data on the areal extent and stocking volume of each
age class within each inventory type. The set of stand-level carbon budgets provides the foundation
for the FCM (see Box 3-2).
    A separate forest carbon model, FORCARB,
developed by the U.S.  Forest Service, is
alternatively linked to TAMM/ATLAS for some of
the analyses (Birdsey 1992a, 1992b). Growing
stock inventories by age class and area are
derived from the ATLAS model and grouped into
248 "management units," defined by region,
owner, species, and site quality. The growing
stock inventories are treated as "snapshots" of
the forest at particular times and converted to
tree (living tree and standing dead trees), soil,
forest floor, and understory carbon. Removals of
growing stock for each management unit are
converted to estimates of harvested carbon.

    The model HARVCARB is  used to examine
the distribution and transfer of post-harvest
carbon quantities predicted by the
TAMM/ATLAS analyses of policy alternatives
(Row and Phelps 1990). The model traces the
flow of forest harvest carbon through five
phases of processing, use, and disposal: from
trees to roundwood logs; through processing
into forest products; into uses  in construction,
manufacturing, and paper products; to
discarded products; and finally, to landfills,
incinerators, and back to the atmosphere.
HARVCARB reports emissions on both a carbon
(i.e., mass) basis and carbon-equivalent (i.e.,
climate impact) basis. The latter result, which
accounts for the generation and release of
methane from landfilled forest products, and the
greater climate impact of methane relative to
carbon dioxide, is used in this report.
HARVCARB was developed at the Institute for
Forest Analysis, Planning, and Policy.

    3.2 AGRICULTURE SECTOR MODELS

    Iowa State University's Center for Agricultural and Rural Development (CARD) has constructed the
Basic Linked System (BLS) of applied general equilibrium models to simulate the world agricultural
                   BOX 3-2
  Estimating Forest Carbon; FCM and FORCARB

     Both the FCM and FORCARB models use the
 volume of merchantable wood (trees that are-farge
 enough to contain salable products} as a basis 1or
 calculating carbon \n the whole stand. Thus, neither
 model calculates the existence of fciomass before
 the stands are old enough to have trees at least B
 inches in diameter in them* Consequently, the
 models show no biomass {or carbon) In stands for
 the ffrst 5 to 15 years after regeneration (depending
 on stand type and region). As a result, biomass
 and carbon are underestimated in the base case
 and scenarios.

     Although this limitation of the model affects the
• calculation of faiomass and carbon in both the
 base case and scenarios, the impacts are likely to
 be important only for some scenarios. As forests
 are managed1 overtime, stands are constantly
 being harvested and regenerated. In the years
 following harvest, or in ihe first few years of new
 stands, non-merchantable volume may become a
 more significant portion of the forest condition, and
 so the carbon models might miss a significant
 amount of carbon. Such situations can arise in the
 first few years following a significant increase In
 harvested acreage, of In the first few years
 following large amounts of new afforestation or
 reforestation* Thust the net carbon accumulation
 for some scenarios, such as afforestation or
 increased biomass energy production, may be
 understated relative to the base case.
                                              13

-------
economy. The BLS consists of 20 national models of the major agricultural producers and 14 regional
models, which account for the remainder of global production. At the international level, all countries
trade in a world market of 10 commodities, which may be aggregated from more extensive lists at the
national and regional level (the U.S. model, for example, contains 23 commodities). For the purposes
of this analysis, only the  model of the United States was used to model the impacts of alternative
agricultural practices and the Conservation Reserve Program (CRP) and the Wetlands Reserve
Program (WRP). Because the BLS was not designed for projections, but rather for determining the
effects of different scenarios, model results are more valid for incremental changes of the scenarios
from the base case than for the absolute values generated in either individual scenario or base case
runs.

    The Resource Adjustment Modeling System (RAMS), was developed by Iowa State University's
Center for Agricultural and Rural Development (CARD) in order to determine optimal  agricultural
production levels for regional Producing Areas (PAs). A Producing Area is an aggregation of counties
(often across state boundaries) that has been defined by the U.S. Water Resources Council as being
a single hydrologicai area. Each PA is sufficiently small that it can also be assumed that production is
homogeneous across the PA. Because each PA acts as an independent decision-making entity, RAMS
allows for region-specific targeting of production practices.

    RAMS is designed as a  short term, static, profit-maximizing, linear programming model of
agricultural production. Profit-maximization is subject to a number of constraints including the
technology available, the resource base,  and  government policy. All activities are characterized by four
dimensions: crop  rotation, tillage, contour management, and irrigation. RAMS  is designed to
accommodate crop and  geophysical process models such as DNDC and CENTURY in order to
provide a more comprehensive evaluation of policies  designed to sequester carbon in agricultural
soils.

    The Denitrification and Decomposition (DNDC)  model, which was developed by the Bruce
Company and the University of New Hampshire, is used to examine carbon and nitrogen dynamics in
agricultural soils. Nitrous oxide (N2O), dinitrogen  (molecular nitrogen) (N2), and carbon dioxide (CO2)
production in soils are evaluated as a function of agriculture practices, soil conditions, and climate
(both precipitation and temperature). Common agriculture practices that are modeled include: tillage
intensity; fertilizer  application; seeding and harvest times; and crop rotations. Rainfall events, soil
moisture, and temperature (determined from the climatic data and soil properties) drive the fluctuating
physical processes. Between rain events, decomposition occurs, which results in CO2 production and
oxidation (including nitrification), while during the rainfall  events themselves the denitrification process
dominates.

    CENTURY was originally developed at Colorado State  University, and has been supplemented by
more recent collaboration with  Michigan State University. CENTURY simulates the dynamics of soil
organic matter (SOM) in both grassland and agricultural  ecosystems. Specifically, the dynamics of
carbon, nitrogen,  phosphorous, and sulphur in soils can be modeled in one month intervals for
periods as long as 100 to 10,000 years; scenarios can be modeled using monthly time steps. SOM in
CENTURY is divided into three categories based on the time scale required for turnover or
decomposition: 1.5 years (active), 50 years (slow), and 1000 years (passive). The decomposition rate
is a function of chemical composition of the soil,  expressed in terms of lignin and nitrogen content,
climatic conditions and location of the organic matter.

    In order to assess both the economic and the physical effects, including changes in  carbon levels,
associated with different agricultural practices, RAMS, the agricultural economics model,  is linked with
DNDC and CENTURY, the soil  carbon models. As a refinement of the PAs used in RAMS, Production
Areas are subdivided into Climate Divisions which allow for a more precise representation of soil and
climate conditions. DNDC and  CENTURY then use geographic information systems (GIS) data to
define land type and area with increased resolution.
                                              14

-------
    ASM, the Agriculture Sector Model of the United States, was developed at the Agricultural
 Economics Department of Texas A&M University to determine changes in U.S. production, consumer
 and producer surplus, prices, international trade, and food processing based on changes in resource
 availability. Production is divided into 63 regions and 36 commodities with region-specific data for crop
 yields and human and natural resource endowments. Each region uses a set of supply and demand
 budgets, which incorporates prices, quantities, and elasticities. Supply budgets  account for
 transportation costs, while demand includes aggregate demand from foreign markets and domestic
 demand in the form of food consumption, Commodity Credit Corporation stock,  feed grains for
 livestock, and food processing. ASM has been modified to include a stumpage supply sector (from
 TAMM), which harvests the trees on both agricultural land and commercial timberland.


 4.  TREE PLANTING SCENARIOS

    U.S. forests currently occupy almost 740 million acres, or about  5 percent of the world's forest
 area, providing an important terrestrial "pool" for carbon (Haynes,  et al.  1994; Birdsey and Heath
 1993). Trees, which are relatively long-lived structures, sequester CO2 from the atmosphere as they
 grow, storing it as carbon in trunks,  limbs, foliage, and roots. Soil  and vegetative cover also provide a
 pool for carbon.  In 1987,  U.S. forest ecosystems stored about 53 billion metric tons of carbon, of
 which over half was in the soil (Birdsey and Heath 1993). Standing (living and dead) trees store
 almost one-third of the forest carbon, and the remainder is stored in the understory or on the forest
 floor (see Figure 4-1). Thus, expanding the carbon pool by increasing the acreage devoted to
 timberlands and other woodlands has the potential to sequester significant additional amounts of
 carbon.
                   Figure 4-1.  Carbon Storage in U.S. Forest Ecosystems by Forest
                             Ecosystem Component (in billion metric tons)
                           Understory O.9
                              (2%)
            Soil 34.O
             (59%) _
                              Forest Ecosystem
                                              Forest Floor 5.3
                                                 (9%)
Foliage 0.4
'(3%)

Other Parts 5.4
(31%)
                                                                       Roots 3.O
                                                                       (17 %)
                                                              Trees
                Source: Birdsey and Heath 1993
    This portion of the report describes the results of the analyses of possible forestation programs.
Section 4.1 examines the timber, carbon and economic impacts of large scale tree planting programs
that are similar in scale to the originally proposed funding levels for the "America the Beautiful"
program. Next, Section 4.2 looks at the implications for both carbon and agricultural production of
expanding existing tree planting programs. Last, Section 4.3 links the forest and agriculture sections
together, in order to capture the interactions between tree planting on  agricultural lands and land
prices. Each of the three sections describes the base case and scenarios evaluated, the models used
for the analysis, and  presents the impacts on carbon and on the forest and agriculture sectors.
                                              15

-------
    4.1 TREE PLANTING ON MARGINAL CROPLAND AND PASTURELAND

    Two key studies (Moulton and Richards 1990, and Parks and Hardie 1992) have identified land
areas in the U.S. that are suitable for establishing forest plantations. These lands are generally
marginal crop and pasture land, which are defined based on soil erosion rates, soil type, and U.S. Soil
Conservation Service land classifications. The carbon sequestration potential of these lands varies
considerably by region and land  quality, reflecting different growth rates and wood densities for
various tree species and climate  conditions (as Table 4-1  illustrates for each region).  Planting on these
lands can generate benefits other than carbon sequestration, including reduced soil erosion,
watershed protection,  conservation of biodiversity, and improved recreation and aesthetic
opportunities.

               Table 4-1
    Average Storage of Carbon In Forest
       Ecosystems by U.S. Region
Region
South Central
Southeast
North Central
Northeast
Rocky Mountain
Pacific Coast
Total U.S.
Carbon Storage
(pounds per acre)
116,748
124,146
162,948
165,021
128,040
205,225
158,225
  Source: Birdsey and Heath 1993
    The current analysis examines the carbon, timber, and economic impacts of federally subsidized
 tree planting programs, using several models (see Section 3). TAMM, an economic model focusing on
 the forest sector and private timberland, is linked to ATLAS, a forest inventory model, which tracks
 forest age-class distribution and productivity classes, and to NAPAP, which models the pulp and
 paper sector. These models are combined with a similar inventory constructed for public lands. For
 public lands, there are no stand-alone inventory models, and inventory assumptions are embedded in
 a variety of harvest planning  models. Estimated future forest inventories are derived using data from
 planning efforts at the level of individual National Forests and other public agencies, such as the
 Forest Planning Model (FORPLAN) used in the National Forests. Carbon data for the forest inventory
 at any point in time are supplied by two alternative forest carbon models, FCM and FORCARB.
 HARVCARB analyses of the flow of post-harvest carbon supplement the results of these two forest
 carbon models. The CGTM, which models global timber markets, provides additional insights into the
 impacts of tree planting on U.S. timber inventories, harvests, and imports and exports.

    This combination of models captures a number of critical impacts and interactions. Together, the
 models provide projections over the  period 1990 to 2040 for carbon stored on forested land, carbon
 stored  in other pools (i.e., in  products produced  using harvested wood or in landfills), impacts on
 timber  inventory, harvest, and other effects. In the analyses, all trees planted and/or retained from
 harvest under a tree planting program are available once they reach minimum harvest age.  Thus,
    5 Minimum harvest age is approximately 20 years in the South, 45 years in the Pacific Coast states, and 70
 years elsewhere.

                                               16

-------
near-term tree planting, by increasing the standing timber available in the future for harvesting, can
significantly affect the prices of both sawtimber and wood-containing products (such as lumber and
furniture) and may affect consumer and producer welfare.

    These models do not, however, capture several important effects. The impacts of federally
subsidized tree planting on stumpage prices will affect not only future decisions about harvesting, but
also decisions made by forest landowners to plant or harvest today, made in response to expectations
about future prices and profits, an effect not captured by the TAMM/ATLAS models (see Box 4-1). In
addition, large-scale tree planting on  agricultural
lands could potentially increase the prices of
crop land. This increase could affect farm
income and/or crop prices,  as well as alter the
cost to the government of planting trees on any
given acreage. The analysis using the ASM
model, which is reported in Section 4.3, partially
captures this interaction. Ongoing work at the
U.S. EPA and U.S. Forest Service continues to
investigate these effects, and so the results
reported here are preliminary. However, they
illustrate the significant potential of expanding
forested land to sequester carbon.
    SCENARIOS AND BASE CASE

    The tree planting analyses  examine the im-
pacts of two alternative funding levels and two
alternative land base/enrollment schedules. For
funding levels, alternative government
expenditures of $110 million and $220 million
annually for ten years are used. These funding
levels correspond to the originally proposed
level of funding for rural tree planting under the
"America the Beautiful" tree planting program,
and to a more aggressive program with double
that funding. The funding covers half the cost of
planting and establishing trees, and 10 years of
land rental costs, or payments  made to owners
of the land where planting occurs.

    Enrollment schedules are designed to forest
land in the most cost-effective manner, i.e., to
enroll first those lands on which carbon could
be sequestered most cheaply.  Because
differences in  the land base can affect the
pattern of cost-effective enrollment, two
alternative schedules of land area (described in
Table 4-2 and Figure 4-2) are used: a low
enrollment schedule and a high enrollment
schedule  (derived from Moulton and Richards
1990, and Parks and Hardie 1992, respectively).
For both enrollment schedules, the majority of acres enrolled are marginal pasture land. These
scenarios assume that, at the end of the 10 year contract period, all acres revert gradually to
agricultural use (less than one-half percent annually) over the time-frame of the analysis. Table 4-2
also presents the enrollment schedule if $110 million in annual funding (for 10 years) is distributed
following the geographic enrollment pattern observed historically in the U.S. over the years 1981-1990.
                  Box 4-1
     investment Behavior in TAMM/ATLAS

    The TAMM/ATLAS models do not Include
price-sensitive land management decisions and,
consequently, do not capture the effects of future
prices on current planting decisions, Bather, the
models project behavior based on historically
observed investment responses, Since there are no
historic data on behavior under conditions explored
fn these scenarios (Len targe-scale tree planting),
the models may not project the most likely
responses (Winnett, et al. 1993), These models
would have a tendency to project little or no
divestment of timber (the observed pattern for
private non-industrial forest owners under
historically observed conditions) even under
conditions of very low prices brought about by
large increases in timber supplies. An analogous
situation would apply to new timber investment
under conditions of very high prices and timber
scarcity.

    In contrast, a model that included responses
to expected prices in management tehavior would
show very different behavioral responses to a large
scale federal tree planting program, in a world of
perfect foresight, owners, in anticipation of lower
future prices when the trees planted today become
available for harvest, may choose to plant less or
accelerate current harvest plans, thus mediating the
impacts of the program on future prices, Simitartyt
higher anticipated; future prices would lead prof tf-
maximTzing owners to delay harvest decisions and
undertake additional near-term planting.

    The true response is bounded by these two
models of behavior, and probably lies somewhere
in between. Results, particularly those for pricesv
could therefore be less  extreme than some of
those presented  here.
                                                17

-------
                                           Table 4-2
                           Regional Distribution of New Forest Under
                        Alternative Enrollment Schedules (million acres)

Region
South Central
Southeast
North Central
Northeast
Rocky Mountain
Pacific Southwest
Pacific Northwest
Total U.S.
$110
low
enrollment
5.8
0
0
0
0
0
0
5.8
Million Annually
(1 0 years)
high
enrollment
6.4
0
0
0.1
0
0
0.5
7.0

historic
enrollment
2.0
1.5
0.3
0.1
0
0
0.3
4.2
$220
Million Annually
(10 years)
low high
enrollment enrollment
8.0
0
0
0
1.8
0
0
9.8
8.5
0.4
2.5
0.2
0
0.1
0.6
12.3
Source: Birdsey and Heath 1993
                   Figure 4-2.  Geographical Regions Corresponding to Enrollment Scenarios
                PACIFIC NORTHWEST
                                                   NORTH CENTRAL
                                                                          NORTHEAST
                                                                                 SOUTHEAST
                     PACIFIC
                   SOUTHWEST
                                                     SOUTH CENTRAL
                Source: Turner, et al. 1993
                                                    18

-------
    The Forest and Rangeland Renewable Resources Planning Act (RPA) of 1974 requires that the
U.S. Forest Service (USFS) of the U.S. Department of Agriculture assess the timber situation and
examine possible future trends in cost and ^availability. USFS performed an assessment in 1989 that
examined the period 1989-2040 (USFS 1990). The base case for the forestation analyses in this report
is a modified version of the 1989 RPA generated by the TAMM/ATLAS system, and reflects knowledge,
as of 1992, about future timber supply, demand, and harvest on private and public lands. The base
case projects a growth in GNP of 2 to 3 percent over the 50 year period, increasing energy costs, and
an increase in U.S. population to 333 million by 2040. Demand for forest products rises following
population  projections, but reflects a slowing of per capita demand. The base case for the analysis
presented here  differs from the 1989 RPA primarily in reducing National Forest timber harvest
projections by approximately 0.5 billion cubic feet per year to reflect revised National Forest
management plans, additional restrictions on log exports, and harvest set-asides attributable to legal
actions under the Endangered Species Act.

    Figure 4-3 presents the range of forest carbon projected by FCM and FORCARB for the
TAMM/ATLAS base case, against which subsequent scenario analyses are compared. The two carbon
models use slightly different assumptions about changes in carbon in the various ecosystem
components (especially soils), reflecting uncertainty in the supporting literature. In particular,
FORCARB assumes that soil carbon levels undergo significant change in response to forest
management activities, with  carbon levels decreasing after harvest and increasing during regrowth. In
contrast, the FCM model assumes that soil carbon levels are relatively insensitive to management
activity and there is little per acre  change in carbon due to harvest and regrowth. Consequently, the
FCM results show a slow  but steady increase in carbon levels, while the FORCARB results show steep
increases in carbon inventories, which are then slowed by the harvest of trees as they reach financial
maturity.
                  Figure 4-3. Total Carbon Storage on U.S. Public and Private Timberland:
                            Range of Model Results for Base Case Scenario
                                         2010
                                                                           2040
                                               Year
                    Sources: Lee 1993c;  Heath 1993
                                             19

-------
    CARBON IMPACTS

    Growth rates change as trees age, so that carbon accumulation rates vary over time. Because
trees grow relatively slowly, the bulk of the carbon is accumulated some years after planting occurs.
Figure 4-4 illustrates the range of results for two of the scenarios — the lower funding, lower acreage
enrollment schedule, and the higher funding, higher acreage enrollment schedule, which together
essentially bound the analysis. Results are also presented for the scenario in which $110 million in
annual funding (for 10 years) is used to plant trees on lands following the historic distribution of lands
planted over the years 1981-1990. The range of carbon accumulation reported for each of these
scenarios also results partly from differences between the two forest carbon models used, FCM and
FORCARB.
                       Figure 4-4. Total Carbon Storage on U.S. Public and Private Timberland:
                                 Increases Relative to Base Case for Tree Planting Scenarios
 1990
   Sources: Turner, et al. 1993; Heath 1993
    As illustrated in Figure 4-4, large scale tree planting has the potential to increase the size of the
forest carbon pool over base case projections by between about 500 million metric tons and 800
million metric tons by the year 2040, with approximately one-fourth of the total achieved by the year
2010. Table 4-3 presents annual carbon accumulation over the base case at decadal intervals. Carbon
results in Table 4-3 are the average of the FORCARB and FCM model results.

    Harvested timber is an  important component of the total forest sector carbon pool,  because the
carbon can be stored not only in forest ecosystems, but also in products manufactured from
harvested trees, and in products discarded  into landfills. The  model HARVCARB examines the
distribution and transfer of  harvested wood  for the TAMM/ATLAS forestation scenarios.  In the base
case analysis, HARVCARB  allocates 81 percent of each year's harvest during the first decade to
storage in  products or landfills. The balance of harvested carbon returns to the atmosphere
                                              20

-------
                                       Table 4-3
  Annual Carbon Accumulation on U.S. Public and Private Timberland Relative to Base Case for
             Three Tree Planting Scenarios, Decadal Intervals (million metric tons)
                                2000
2010
2020
2030
$220 million annually for 10
years, high enrollment
$110 million annually for 10
years, low enrollment
$110 million annually for 10
years, historic distribution
6.8
4.1
3.5
16.7
9.6
6.5
19.4
10.8
7.5
13.0
6.2
5.5
 Note: Forest carbon estimates are presented as the average of FORCARB and FCM results.
 Source: Turner, et al. 1993; Heath 1993
immediately, primarily as a result of energy production.  In succeeding years, carbon from current
harvests continues to be allocated largely to products, and carbon moves from products to both
disposal pools and the atmosphere as older products are retired.

    Because a large portion of the carbon contained in harvested timber remains in products for some
length of time, these products represent an additional store of carbon, over and above that stored on
forested lands. The forestation scenarios show an average annual carbon accumulation (post-harvest)
in products of 1.7 to 2.4 million metric tons over the 50 year period. However, the potential climate
impact of this carbon accumulation may be offset when it is adjusted to account for the effect of
methane emissions from landfilled forest products (Turner, et  al.  1993).

    TIMBER INVENTORY EFFECTS

    The scenario analyses suggest that tree planting programs funded at $110 million to $220 million
per year for ten years would lead to large increases in U.S. timberland area. Results suggest that the
tree planting scenarios would have the largest impact on privately owned U.S. softwood growing stock
inventories and harvest. These effects are greatest in the South Central region where a majority of the
land enrollment is projected to occur under all scenarios. As Figure 4-5 illustrates, U.S. softwood
inventories on privately owned lands in the scenarios increase over the base case by about 2 percent
in the year 2010 and by between 6 and 10 percent in the year 2040.7 Table 4-4 illustrates softwood
harvest rates, for selected years and forestation scenarios.

    ECONOMIC AND TRADE EFFECTS

    Projected increases in timberland are likely to have substantial impacts  on U.S. timber markets,
particularly in the South Central region where the largest increases occur. The analysis shows that
increases in timber supplies drive down stumpage prices relative to the base case, particularly
softwood stumpage prices in the South, where prices become low and steady toward the end of the
     To the extent that it substitutes for energy from fossil fuel use, the use of forest products can be considered to
be a permanent carbon sink, or an energy offset.

     Timber inventory as used here is the growing stock volume, defined as the cubic volume of all live trees of
commercial species, meeting specified standards of quality or vigor, which are at least 5.0 inches in diameter at
breast height.

                                               21

-------
              Figure 4-5.  U.S. Softwood Inventory Projections For Privately Owned
                         Timberland: Base Case and Tree Planting Scenarios
          250 -r
          240 - -
          230 --
          220 --
          210 --
          200 - -
          180
                                                       $110 M Hllon Low
                                                       E n rollm e nt
                                                  	$220 M Hllon. riign
                                                       Enrollm ent
                                                  —	$110 Million Historic
                                                       D 'stributio n
            1990
            Source: Haynes. et al. 1994
                                     Table 4-4
           U.S. Softwood Harvests for Tree Planting Scenarios: All Owners
                                 (billion cubic feet)
$110 Million Annually $220 Million Annually
(10 years) (10 years)
Year
1990
2000
2020
2040
Base Case
11.0
11.8
14.1
15.3
Historic
Distribution
11.0
11.8
14.1
15.6
Low
Enrollment
11.0
11.8
14.2
15.8
High
Enrollment
11.0
11.8
14.3
16.0
 Source: Haynes, et al. 1994
time-frame examined.8 If private owners plant at the levels assumed under these scenarios,
reductions in stumpage prices are likely to be tempered by actions taken by other landowners to plant
less area to trees, an  effect which is the subject of ongoing investigation by the U.S. EPA. To the
extent that prices decline, consumers would tend to benefit and owners of timber would lose.
   8 Stumpage price is paid for standing timber in the forest, and is typically measured in dollars per thousand
board feet.
                                                 22

-------
    Although softwood inventory and harvest projections increase under each tree planting scenario,
stumpage price projections fall relative to the base case (see Tables 4-5 and 4-6). The price decline is
greatest in the South, where the majority of marginal crop and pasture lands are situated, and, hence,
where planting occurs. Prices in the South become low and steady soon after the timber inventory
begins to increase, i.e., when the wood becomes merchantable (after 2 or 3 decades). Softwood
stumpage prices in the Rocky Mountains and Pacific Coast show slight declines, relative to the base
case, over the time frame of the analysis, while prices in the North are largely unaffected, as are
hardwood sawtimber stumpage prices. The declines are greatest for the largest tree planting program.
                                          Table 4-5
                                Stumpage Prices in the Base Case
                                 (dollars per thousand board feet)
                                 2000
            2010
            2020
                                                                       2030
                                                   2040
  Softwood Stumpage Prices

     North

     South

     Rocky Mountains

     Pacific Coast
 50

180

102

247
  Hardwood Sawtimber Stumpage Prices

     United States                   463
 66

260

236

311



559
 88

306

287

365



665
101

306

259

367



784
111

297

279

358



903
 Source: Haynes, et al. 1994
    As U.S. stumpage prices decline, U.S. forest products become more competitive, and Canadian
lumber imports to northern U.S. markets decline after 2010, falling to between 7 and 9 billion board
feet in 2040, as compared to 13 billion board feet in the base case projection (Haynes, et al. 1994).
The CGTM analysis finds similar trade impacts. In addition, because there is no or little enrollment of
lands in these programs in the Pacific Northwest, exports from this region to Pacific Rim  markets are
not affected (Perez-Garcia 1994).

    4.2 CONSERVATION RESERVE AND WETLANDS RESERVE PROGRAMS

    The analyses of the Conservation Reserve Program (CRP) and Wetlands Reserve Program (WRP)
provide another view of the impacts of government subsidized tree planting. CRP was established as
part of the Food Security Act of 1985 to foster the withdrawal from agricultural production of between
40 and 45 million acres of highly erodible grasslands and wetlands. To be eligible for inclusion in the
CRP,  land must have an average erosion rate of at least 19.1 tons of soil per acre — nearly three times
the national average. By this criterion, an estimated 24 percent of all U.S. cropland — mostly in the
Corn  Belt, northern and southern plains, and mountain  regions — is potentially eligible. Without special
dispensation, however, CRP signups are limited to 25 percent of a county's cropland, in  an  attempt to
spare local economies significant damage. This limitation reduces eligible land by almost one-third, to
just under 70 million acres (Trexler 1991). To encourage landowners to participate in the CRP, the
government shares the cost of converting the land into  alternative uses (such as trees and other
cover) and provides rental payments throughout the 10-year CRP contracts. Currently, there are
approximately 36 million  acres enrolled in the program,  about 2.4 million of which are planted with
trees, and the remainder of which are in grasses.
                                              23

-------
                                          Table 4-6
                          Stumpage Prices In the Tree Planting Scenarios
                                (dollars per thousand board feet)
2000
$110 million annual funding for 10 years,
Softwood Stumpage Prices
North 50
South 178
Rocky Mountains 102
Pacific Coast 248
Hardwood Sawtlmber Stumpage Prices
United States 463
$220 million annual funding for 10 years,
Softwood Stumpage Prices
North 50
South 178
Rocky Mountains 102
Pacific Coast 248
Hardwood Sawtimber Stumpage Prices
United States 463
$110 million annual funding for 10 years,
Softwood Stumpage Prices
North 50
South 179
Rocky Mountains 102
Pacific Coast 244
Hardwood Sawtimber Stumpage Prices
United States 463
2010
low enrollment

66
237
239
310

559
high enrollment

64
228
239
309

558
2020


88
235
265
342

664


82
199
260
334

663
2030


101
33
230
314

782


84
*
210
293

779
2040


111
*
217
280

899


90
*
195
247

896
historic distribution

66
250
243
316

559

88
256
270
351

664

100
203
252
334

782

110
211
253
313

900
 Note: "*" indicates prices outside the range of model results.
 Source: Haynes, et al. 1994
    The Wetlands Reserve Program was established as part of the 1990 Farm Bill to restore and
protect wetlands that have been partially or fully converted to agricultural use. The WRP places
accepted land areas under 30 year or permanent easements that prohibit the draining of wetlands.
The WRP encourages land owners to enter the program by paying compensation based on the type
of easement and the appraised land value, up to the full value of the land. The program also pays for
the cost of any restoration activities. The program goal was to have enrolled slightly less than 1 million
acres by 1995, although projected enrollment is a fraction of this amount.
                                               24

-------
     Because the reserve programs operate on agricultural land, the BLS model provides the results for
 crop, income, and price effects in the agriculture sector. Carbon impacts are calculated from the BLS
 results using the FCM. Together, these analyses provide results that encompass both carbon and
 economic impacts. Although the  impacts are based on specific policy options, they also illustrate the
 impacts that other large scale tree planting programs, such as those described in Section 4.1, could
 have on agriculture.

    SCENARIOS AND BASE CASE

    The analysis examines the carbon and economic impacts of tree planting for two alternative CRP
 scenarios and a targeted WRP scenario. For the CRP analyses, the base case assumes that the
 original CRP target enrollment of 40 million acres is achieved by 1995, then declines over time to 17.5
 million acres, an outcome that is  considered to be likely after current contracts expire. The CRP
 scenarios reflect different assumptions about the size of future programs and alternative uses of CRP
 land. Two alternative proposals — 40 million acres and 50 million acres — are compared to the base
 case, as illustrated in Figure 4-6.  Historical signups under the CRP guide enrollment patterns under
 the base case and CRP scenarios. For both the base case and the 40 million acre CRP scenario, 3.7
 million acres are planted with trees and the remainder is grassland. The incremental 10 million acres
 in the second CRP scenario is  planted with trees.
                          Figure 4-6. Annual Acreage Under CRP Scenarios
                50 - •

                45 •-

                40 - -

                35 - ;

                30 ••

                25 • •

                20 •-

                15 - •

                10 -•

                5 --
                0
                 1990
                         1995
                                 2000
                                         2005
                                                 2010
                                                 Year
                                                         2015
                                                                 2020
2025
        2030
                 Note: In the base case, the current program reaches 40 million acres in 1995 and
                 then contracts until 17.5 million acres remain in 2005.  CRP-40 retains all 40 million
                 acres after 1995. CRP-50 expands the program to cover 50 million acres by 2005.

                 Source: Reese, et al. 1993
    The third scenario, WRP, simulates the introduction of a 5 million acre reserve of wetlands
consisting predominantly of drained bottomland previously planted to agricultural crops. These acres
are targeted to establish hardwood tree stands for the potential carbon benefits of the trees, and to
provide wetlands benefits such as flood control, habitat preservation, and improved water quality. The
WRP also involves considerably different crops than the CRP, as displayed in Figure 4-7.
                                               25

-------
                              Figure 4-7.  Composition of Reserve Scenarios
                  Typical CRP
                                                                    WRP
         Other
          24%
12%
Wheat
 30%
                      Other
                       11%
Wheat
 11%
                Other Grains
                   18%
                                    Corn
                                    12%
             Soybeans
               42%
Other Grains
    6%
 Source: Reese, et al. 1993
    CARBON IMPACTS

    The FCM model suggests that expanding or extending the CRP program and using the WRP to
plant hardwood trees can sequester significant additional carbon dioxide. As Figure 4-8 illustrates,
maintaining the full 40 million acres results in an additional carbon accumulation of almost 200 million
metric tons in 2015 and nearly 350 million metric tons by 2035. This accumulation is relative to the
base case; full implementation of the current CRP plans (assuming that land begins to revert to
agricultural production after 1995). Carbon storage in the base case is approximately 1,050 million
metric tons in 1995, growing to 1,160 million metric tons in 2015 and 1,360 million metric tons in 2035.

    Expanding the program to 50 million acres, by planting an incremental  10 million acres with trees,
more than doubles the carbon accumulated on CRP lands, relative to the base  case. Under full imple-
mentation of the 50 million acre CRP program, the total carbon pool on CRP lands, including soil,
trees, understory and other forest ecosystem components, is estimated to reach almost 2,500 million
metric tons, a gain in excess of 1,100 million metric  tons over the base case in the year 2035. The
smaller WRP program results  in almost 200 additional million metric tons of carbon by the year 2035.
For the three scenarios, Table 4-7 presents annual carbon accumulation relative to the base case.

    IMPACTS ON AGRICULTURE

    Maintaining or expanding  the reserve programs can affect agricultural output by removing
agricultural land from production  and driving up commodity prices. The BLS model is used to estimate
the economic impacts of the CRP and WRP scenarios on the agriculture sector of the United States.
Crop acreages in both CRP scenarios are consistently lower over time than in the base case (see
Table 4-8). In percentage terms, soybean and rice acreages are the least affected of the major crops,
and planted acres of the "other grains" are reduced the most, with barley acres falling most
dramatically. For most crops the  reductions  in acreage are slightly offset by increases in yield on
planted acres, as producers farm their remaining acres more  intensively in  response to higher output
prices. On balance, production of all major crops is less under the CRP scenario than the base case,
and producer prices are concomitantly higher (see Table 4-8).
                                              26

-------
                            Figure 4-8.  Carbon Storage Under Reserve Program
                                       Scenarios: Increases Relative to Base Case
          b.
          c   600
                  Note: CRP estimates are relative to CRP base case. WRP estimates are relative to carbon
                  storage in 1995 on WRP lands that will be planted with hardwood trees.


                  Sources: Lee 1993a; Lee 1993b
                        Table 4-7
      Annual Carbon Accumulation for CRP and WRP
             Scenarios Relative to Base Case,
           Decadal Intervals (million metric tons)
                    2000
2010
2020
                                                 2030
CRP-40
CRP-50
WRP
11.6
31.5
2.2
7.9
15.9
4.4
7.7
34.5
7.0
7.2
31.6
4.6
 Note: WRP estimates are relative to flux in 1995 on WRP lands that
 will be planted with hardwood trees. Forest carbon estimates are
 derived from the FCM model.
 Source: Lee 1993a; Lee 1993b
    Because the output of feed grains is significantly affected, feed prices are higher under the CRP
scenarios, and livestock production is generally lower under the scenarios, reflecting higher feed
costs. Beef production increases in both CRP scenarios relative to the base case (by 1  percent to 2
percent) and production of pork, milk, poultry,  and eggs falls. Pork shows the largest decline over the
time period, on the order of 3 percent to 4 percent, while declines for milk, poultry, and eggs are
generally less than one percent (Reese, et al. 1993).
                                                 27

-------
                                             Table 4-8
        Impacts of Reserve Program Scenarios on Agricultural Production and Producer Prices:
                      Percent Change Relative to Base Case, Decadal Intervals

                                   CRP - 40 MILLION ACRES PLANTED
Year
Acree Harvested
2000
2010
2020
3030
Producer Prices
2000
2010
2020
2030
Wheat

*7
-8
*$
-7

4
4
4
6
Rice

-2
-2
4
-2

2
1
1
3
Corn

*4
-4
-4
-5

5
6
§
5
Other Grain

""- „**
-15
-if
-18

"+ •. •. V.W.
-
4+
'..
-
Soybean

»1
-2
•«
1

4
3
$
4
Cotton

«e
-7
*7
-7

1
1
1,
0
                                   CRP - 50 MILLION ACRES PLANTED
Year
Acres Harvested
2000
2010
2020
2030
Producer Prices
2000
2010
2020
2030
Wheat

-8
-10
-10
-9

$
5
5
7
Rice

-a
-2
-2
-2

2
1
1
3
Corn

„„;! 	
-6
-S
-7

7
8
$
6
Other Grain Soybean

•1K -3
-20 -5
•S.9 -5
-28 -1

a
6
— ^ S
8
Cotton

-7
-9
•8 '
-10

2
1
2
-0.3
                                    WRP - 5 MILLION ACRES PLANTED
Year
Acres Harvested
2000
2010
2020
2030
Producer Prices
2000
2010
20?0
2030
Wheat

0
0
6
0

1
1
1
1
Rice

Q
0
0
0

*
*
#
1
Corn

0
-2
-1
-2

1
1
1
1
Other Grain Soybean

0 -2
-1 -4
'-1 " " " -4
-1 -3

2
3
S
4
Cotton

*
*
*
*

*
*
*•
*
Note: "*" indicates an increase or decrease of less than 0.5%.
Source: Reese, et al. 1993
                                                 28

-------
    The WRP scenario has similar, but less dramatic, effects. The production of soybeans and corn is
most affected, with acreage reductions of about 4 percent and 2 percent, respectively. In contrast,
acreage falls by at most 1 percent for other crops. As a result, soybean producer prices increase by
about 3 percent and corn prices increase by about 1 percent (see Table 4-8). Correspondingly small
impacts occur in livestock; production of pork and of poultry falls by 1 to 2 percent, while production
of other livestock products is relatively unaffected (Reese, et al. 1993).

    ECONOMIC IMPACTS AND GOVERNMENT COSTS

    Under the CRP scenarios, per capita consumption of most commodities changes only slightly,
although consumption of grain products and most meat fall  by 1 to 2.5 percent after the programs
have been fully implemented. As a result of these changes,  producer net returns increase for crop
producers, but decline significantly for livestock producers. Government price support payments to
crop producers fall by more than the cost of the CRP programs.9 Combined returns in the agriculture
and livestock industries, i.e., the net result of changes in income and in government payments, are
about $350 million to $450 million per year lower under the CRP scenarios. By  comparison, total net
U.S. farm income between 1989 and 1991 averaged $47 billion per year. Consumers are generally
worse off with  slightly lower consumption and higher retail prices,  resulting in about a 2 percent
increase in total  consumer expenditures (Reese, et al. 1993). The effects of the WRP are similar;
results for all three scenarios are reported in Figure 4-9, for selected years.

    4.3 THE AGRICULTURE  SECTOR: IMPLICATIONS FOR FEDERALLY FUNDED TREE PLANTING

    Large-scale tree planting affects not only the forest sector but also the agriculture sector.
Subsidized tree planting takes marginal land out of agricultural production and  may raise crop prices,
with consequent welfare impacts for consumers and producers in the agriculture sector, as discussed
in Section 4.2. Taking land out of production may also increase land prices, which in turn increases
the subsidy that is necessary to induce landowners and others to plant trees. Thus, the cost to the
government of funding tree planting on a given number of acres can be higher when  agricultural price
effects are considered. The cost per ton of carbon sequestered will also be higher. Agricultural price
effects may also lead to welfare changes for producers and  consumers in both the agriculture and
forest sectors.

    To assess the carbon  and economic implications of land price increases for the agriculture sector,
the U.S. EPA initiated an additional series of tree planting analyses. The analyses reported in this
section complement the analyses reported in Section 4.1 and Section 4.2. These analyses use the
ASM, a U.S. Agriculture Sector Model (Adams, et al. 1993; and Chang and McCarl 1991). In the ASM
analyses, land currently enrolled  in the CRP is released at the termination of the program and either
reverts to cropland or is available for tree planting. For the analyses, TAMM provides regional forest
sector supply and demand curves and other information about the demand for wood  products.
Estimates of the cost of establishing trees are based on estimates in Moulton and Richards (1990),
and carbon yield data are taken from Richards (1992).

    SCENARIOS

    The ASM analysis evaluates the welfare and  carbon impacts of tree planting programs under
several alternative assumptions. First, the analysis examines  the implications of using alternative land
bases for planting trees: cropland, pastureland, and wetlands. These land types differ in carbon
sequestration potential, in  distribution within the geographic  United States, and in current agricultural
     Program costs are not handled endogenously in the models; rather they are based on estimates by analysts
familiar with the Conservation Reserve Program. It Is estimated that additions to the CRP in the Great Plains states
would cost 40 to 50 percent of the current $49 to 50 per acre. Environmentally sensitive acres added to the CRP
are expected to cost $70 to 80 per acre.

                                              29

-------
       Figure 4-9.  Impacts on Agriculture for CRP and WRP
                  Scenarios, Change from Base Case

               CRP Maintained at 40 Million Acres
     I
    I
              2010
                          2020
                                      2030
               CRP Expanded to 50 Million Acres
     JRJ

     1
               2010        2020        2030

               WRP Established on 5 Million Acres
        -3000
        -4000 J  2010
                           2020
                                       2030
  mi  Change in net farm income
  INI  from crop production

D      Change in government transfer
      payments for program crops
•   Change in net farm income
   from livestock  production

rca Change in government payments
tia for reserve programs
Source: Reese, et al. 1993
                                   30

-------
use. Using alternative land bases, therefore, produces different price impacts for land, crops, and
livestock, and different impacts on economic welfare and carbon sequestration. The three specific land
scenarios reported here involve planting 6.3 million acres on cropland, 7.5 million acres on
pastureland, or 4.6 million acres on wetlands. These programs are roughly similar in magnitude to the
smaller tree planting program in Section 4.1 (low enrollment and $110 million in funding) and the
wetlands program described in Section 4.2, although the land bases are not directly comparable.

    Second, several different enrollment patterns are analyzed. Because land varies in productivity,
land price, and the cost of establishing trees, the cost of sequestering a ton of carbon varies within
each of the land types.10 A program that enrolls lands in order of carbon sequestration  costs will
sequester a given amount of carbon at the lowest cost. For most analyses reported here, land is
enrolled on a least-cost basis: for a given land class,  acres are planted to minimize the cost per ton of
sequestering carbon, so that lands that can sequester carbon most cheaply are enrolled first. For
cropland,  however, an alternative  pattern of enrollment is also analyzed, in which the regional
distribution of planted lands follows the pattern of lands  enrolled in the current Conservation Reserve
Program.

    Last, the analysis examines two alternative harvesting assumptions: optional harvests and no
harvests. In the optional harvest scenario, the ASM simulates the decision faced by farmers in each
region to plant or harvest trees based on economic considerations. While allowing harvesting reduces
the amount of carbon stored on planted lands,  it provides  an additional incentive for landowners to
plant trees, an effect that may offset the potential carbon loss from harvesting, as described more fully
below.

    The ASM has several advantages. Because it is a detailed model of the agriculture sector, it
directly models impacts on land prices (previous analyses  have assumed no impacts on these prices),
captures farm program effects, and reports welfare and other impacts  in the farm sector. The ASM
analyses are, however, static; all planting essentially takes  place in a single time period, and so the
model reports average carbon per acre per year, or average cost per ton per year,  rather than time
paths of carbon, timber, acres, and costs. While this approach captures average carbon accumulation
over the rotation of the trees that are planted, it tends to overstate carbon accumulation that would
occur during the early years of a multiple year tree planting program and to understate  accumulation
during later years.

    IMPACTS ON CARBON, COSTS, AND WELFARE

    Figure 4-10 presents average  annual carbon accumulation for three land types, under alternative
assumptions about harvesting. In  this figure, the highest average annual sequestration rate under the
no-harvest assumption occurs for cropland, followed closely by pastureland, and then by wetlands.
Although less cropland is enrolled than pastureland, higher annual carbon accumulation on cropland
reflects the higher productivity of these lands, at the assumed levels of planting.

    The relative productivity  of these lands  also influences  the costs of sequestering carbon, which are
reported in Table 4-9. In this table, average cost is calculated as the annual cost (to the government)
of the tree planting program divided by the amount of carbon sequestered annually. In turn,
government costs represent the subsidy that the government would have to pay landowners to induce
them to plant the additional trees. This subsidy, therefore, will be higher if land rents are higher.  On a
per ton basis, the subsidy will  be lower if the enrolled land is more productive. The  cost per annual
   10 Cost, in this analysis, is calculated as the full "opportunity cost" of planting trees, i.e., the amount farmers
must be paid to induce them to plant trees, taking into account land prices and profitability of the land in alternative
uses, the cost of establishing trees on the land, and revenues from harvesting mature trees.  Cost is not, therefore,
fully comparable with costs reported in Section 4.1, where it was assumed that the government fully compensates
farmers for foregone land rent, and pays half of the cost of establishing trees, regardless of whether these sum to
more or less than the opportunity cost  of the land.

                                              31

-------
                         Figure 4-10.  Average Annual Net Carbon Accumulation:
                                    Tree Planting on Three Land Types
                        Crcpland             Pastureland


                Source: Callaway and McCarl 1994
ton of carbon, under the assumption of no harvests, is lowest for wetlands, followed closely by
cropland and more distantly by pastureland.

    While the results for the restricted harvest scenario illustrate carbon and cost impacts most clearly,
a more realistic assumption is that harvesting occurs. Allowing harvesting will have several competing
effects. First, harvesting generates some income for those who plant; a lower subsidy, therefore,  is
required to induce additional  planting. In addition, harvesting forested land may lead to more land
                                          Table 4-9
                         Impacts of Tree Planting on Different Land Bases
Cropland

Annual Government Cost
($ millions)
Annual Carbon
(million tons)
Annual Average Cost
($ per ton)
No
Harvest
100
14.1
6.90
Optional
Harvest
60
11.7
5.00
Pastureland
No
Harvest
220
13.9
15.70
Optional
Harvest
90
14.0
6.60
Wetlands
No
Harvest
70
11.1
6.30
Optional
Harvest
40
8.9
4.40
  Source: Callaway and McCarl 1994
                                                32

-------
returning to agricultural use, mitigating the impacts of tree planting on land prices and, thus, reducing
the costs to the government of subsidizing forestation. At the same time, harvesting a large amount of
timber would tend not only to depress stumpage prices significantly, but partially to reduce carbon
sequestration gains.

    As displayed in Figure 4-10, allowing harvesting reduces the carbon stored on cropland and
wetlands,  relative to a situation of no-harvest. Government costs drop substantially, however, as
illustrated in Table 4-9. On balance, the cost-per-ton of carbon sequestered drops significantly for all
three types of land when harvesting can occur.

    Both the amount of carbon sequestered  per acre, and the cost of sequestration per ton of carbon,
are highly sensitive to the pattern of enrollment. When lands are enrolled on a least-carbon-cost basis,
costs of sequestering  a given amount of carbon will be lowest. Alternatively, 6.3 million acres of
cropland could be enrolled in a pattern reflecting the regional distribution of lands historically enrolled
in the CRP. As illustrated in Table 4-10, the two patterns of enrollment produce dramatically different
results. Although average annual carbon accumulation is slightly higher under the historic enrollment
pattern, costs more than double. When harvests are restricted, the costs of sequestering  carbon using
the historic distribution are more than triple those using a least-cost enrollment pattern.
                                           Table 4-10
               Impacts of Tree Planting on Cropland with Varying Enrollment Strategies
                                  Least-Cost Enrollment

                             No Harvest     Optional Harvest
                    Historic Enrollment

              No Harvest     Optional Harvest
  Annual Government Cost         100
  ($ millions)

  Annual Carbon                  14.1
  (million tons)

  Annual Average Cost             6.90
  ($ per ton)
 60


11.7


5.00
 340


 15.7


21.60
 150


 12.6


11.61
  Source: Callaway and McCarl 1994
    Finally, the ASM provides measures of the change in economic welfare and prices in the
agriculture and forest sectors. The results for the agriculture sector, displayed in Tables 4-11 and 4-12,
are similar to those in Section 4.2. Crop prices (which do not include farm program payments) and
livestock prices tend to rise slightly as a result of tree planting on agricultural lands, which results in
declines in consumer surplus.  These results are particularly pronounced for the historic enrollment
pattern, which plants more trees in regions that are important to crop production. Agricultural
producers show a decrease in welfare when land is enrolled in order of carbon cost, and an increase
in welfare under the historic enrollment scenario. As in the analysis in Section 4.2, impacts under
    11 Table 4-11 reports Fisher price indices for agricultural crops (raw agricultural crop prices not including farm
 program payments); livestock; and forest consumer prices for logs (which gives an index of prices to all forest
 producers at the mill dock).
                                                33

-------
either scenario, while apparently large in dollar terms, are relatively small as a percent of the base
case surplus for agricultural producers.12


                                          Table 4-11
           Economic Welfare Effects of Federal Subsidies for Tree Planting on Cropland
                                (million dollars, percent change)
                                     Least-Cost Enrollment
                                   Historic Enrollment
                                   No Harvest
                 Optional
                 Harvest
                        No Harvest
                                Optional
                                Harvest
  Forest Sector
  Consumer Surplus
  Producer Surplus
  0
-10
(0%)
(-1%)
 190  (2%)
-210  (-13%)
            0  (0%)
            0  (0%)
               230   (2%)
              -240   (-15%)
  Agriculture Sector
  Consumer Surplus
  Producer Surplus
 140
-100
(*)
(*)
 -10
 -10
(*)
(*)
-250
 110
(*)
(*)
-610
 330
(*)
(1%)
  Notes: Surplus changes estimated relative to surplus in a base case with no planting subsidies.
  "*' Indicates an increase or decrease of less than 0.5%.
  Source: Callaway and McCarl 1994
                                             Table 4-12
                    Price Effects of Federal Subsidies for Tree Planting on Cropland:
                                         Fisher Price Indices
                                    Least-Cost Enrollment

                  Baseline     No Harvest     Optional Harvest
                                      Historic Enrollment

                                No Harvest     Optional Harvest
Crops
Livestock
Logs
100
100
100
99.9
100
100
100.1
100
88.6
100.7
100.2
100.0
101.5
100.4
87.1
 Source: Callaway and McCarl 1994
    Economic impacts in the forest sector as a result of tree planting are significantly affected by
whether or not harvesting occurs. When harvests are restricted, trees that are planted do not
ultimately enter timber markets, and so there is no change in the log prices or in the welfare of
producers and consumers, as displayed in Tables 4-11 and 4-12. When harvests are allowed to occur,
tree planting reduces forest product prices significantly, resulting in benefits to consumers and
adverse impacts on producers.
    12 Differences in the magnitude and direction of results between Sections 4.2 and 4.3 may reflect differences in
the number of acres enrolled, in enrollment patterns and the assumptions about the rates at which land enrolled in
the CRP reverts to alternative uses or become available for tree planting, in how the farm program is modelled, and
In interactions between the agriculture sector and the forest sector (which is absent in the CRP analysis) in the
ASM.
                                                 34

-------
5.  OTHER FOREST POLICY SCENARIOS
    Policies that increase the acreage of standing trees by reducing harvest can expand the amount
of carbon stored on forest land. Two scenarios, reduced harvest in National Forests and increased
rates of utilization of recycled fiber (which decreases the demand for virgin fiber), provide information
on the carbon, timber, and economic impacts of such forest policies. A third scenario, increased use
of wood fuels, which reduces carbon emissions by displacing fossil fuel generated energy, is also
examined. Finally, this part of the report examines the implications of combining policies to take
advantage of the potentially offsetting economic impacts.

    5.1  INCREASED RATES OF RECYCLED PAPER UTILIZATION

    Policies designed to increase the use of recycled paper and fiber in the paper and pulp making
process (see Box 5-1) increase the size of the forest carbon pool in three ways. First, manufacturers of
pulp and paper products that use greater amounts of recycled fiber use less virgin (freshly harvested)
fiber, reducing the harvest of pulpwood-sized trees. By decreasing demand for harvested wood,
increased use of recycled fiber can increase forest inventories and enlarge the forest carbon pool.
Second, trees left uncut grow larger over time, further increasing forest and carbon inventories. Last,
to the extent that the wastepaper in landfills eventually decays and returns greenhouse gases to the
atmosphere, increased utilization of recycled fiber can reduce emissions associated with the forest
products pool. Pulpwood-sized trees that grow into sawtimber sizes can be cut for long-lived solid
wood products, which sequester carbon for longer periods of time than do paper products.
    Although increased recycling has the
potential to expand forest carbon, timber market
effects may mitigate the beneficial impacts of
recycling on forest carbon. If decreased demand
for harvested timber for paper production
reduces long-term prices in the forest sector,
then less near-term planting may occur, in
anticipation of lower returns in the long term. At
current rates of recycling, however, projections
for the U.S. suggest that serious supply
shortages will exist for sawtimber in the future,
resulting in rapidly rising stumpage prices.
Increased recycling, therefore, can extend fiber
supply and partially mitigate potential increases
in stumpage prices over time;  the scenario
results presented below suggest that, with
increased recycling, real prices of softwood
remain relatively stable (rather than rise) through
the time period.

    This analysis employs the  same models
used for the analyses of large scale tree
planting presented in Section 4.1. The NAPAP
model provides the framework necessary to
analyze the impacts of policies designed to
increase recycling rates. This model  (described
in Section 3) simulates the evolution of process
technology (for using recycled fiber and virgin
wood fiber) for all primary paper and paperboard products (such as newsprint, printing and writing
paper, and linerboard). The model combines regional information on supply and demand,
manufacturing processes, and transportation costs to project future technological changes,
production, capacity,  imports and exports, and related market equilibria for the United States and
                 Box 5-1
         Paper Recycling in the U,S.

    The U.S. currently recycles severed types of
paper, including newspaper, and the rate at which
recyclable paper is being utilized has risen from
around 25 percent fn the late 1980s to almost 30
percent fn 1992, Continuing growth In recycling
rates is technically feasible, particularly for
newspaper, and a variety of policies are being
considered at both the federal and State levels to
increase the rate at which wastepaper Is used.

    Reducing the pressure on landfills provides
much of the impetus for recycling programs; in
1990, paper and paperboard comprised
approximately 37.5 percent, by weight, of the
material deposited in landfills. Policies that have
been proposed to improve waste paper recycling
include minimum recycled content standards for
newspaper, per-unit disposal fees fortrasrt
removal, mandatory recycling rylesf and financial
incentives to use or recycle wastepaper, increased
recycling may also have added benefits for carbon
storage,

Source; Tnce 1994
                                               35

-------
Canadian pulp and paper sector. Combining this model with TAMM/ATLAS provides the necessary
link for analyzing the implications of increased paper recycling on forest product markets and on
carbon storage. The CGTM model provides additional insight into the regional distribution of impacts
on pulpwood and sawtimber markets within the U.S., and the associated effects on international trade
in the forest products sector.

    SCENARIOS AND BASE CASE

    Increasing the use of recycled fiber in the manufacture of paper and board replaces a percentage
of the newly harvested (virgin) fiber in the production process. Using less virgin fiber reduces the
harvest of trees and leaves more of the forest, intact to grow longer and to sequester more  carbon.
The increased recycling scenario differs from the base case both in how quickly higher levels of
recycled  fiber use are attained, and in the magnitude of the final level of recycled fiber use.  The
increased recycling scenario achieves 45 percent utilization by 2000 and maintains that level through
2040. This scenario, which was derived from suggestions by the American  Paper Institute, assumes
that policies and programs are aimed at reducing volumes of waste (through more efficient  product
use, public education, disposal fees,  and other means).
                   Figure 5-1. Total Carbon Storage on U.S. Public and Private Timberland:
                             Increases Relative to Base Case for the Increased Recycling Scenario
            1990
                          2000
                                        2010
                                                      2020
                                                                    2030
                                                                                  2040
                                               Year
               Sources: Turner, et al. 1993; Birdsey and Heath 1993
    CARBON IMPACTS

    Increasing the use of recycled fiber increases the supply of standing timber and the associated
carbon storage, as Figure 5-1 illustrates. Carbon storage above base case levels rises to between 400
and 600 million metric tons by the year 2040, a level that is comparable to that achieved by the large
scale tree planting programs discussed in Section 4.1 (the range reflects estimates from FORCARB
and from FCM). Because the gains in carbon storage under the increased recycling scenario depend
less on new growth than on conserving existing standing stock, increases in carbon occur more
rapidly than in the tree planting scenarios; more than half of the carbon accumulates by the year
                                              36

-------
2010. Table 5-1 presents annual carbon accumulation at decadal intervals. The HARVCARB analysis
suggests additional average annual carbon accumulation in post-harvest pools of 1.9 million metric
tons of carbon per year, over the 50 years of the analysis. However, the potential climate impact
benefit of post-harvest carbon accumulation may be offset when the relative climate impact of
methane is taken into account (Turner,  et al. 1993).
                                      Table 5-1
  Annual Carbon Accumulation on U.S. Public and Private Timberland Relative to Base Case
          for Increased Recycling Scenario, Decadal Intervals (million metric tons)
                             2000
                                            2010
                                                           2020
                                                                          2030
  Increased Recycling
13.5
15.6
                             10.0
                                                                          4.4
  Note: Forest carbon estimates are presented as an average of FORCARB and FCM results.
  Source: Turner, et al. 1993; Heath 1993
    TIMBER INVENTORY EFFECTS

    As the demand for virgin fiber declines, fewer acres are harvested and timber inventories rise (see
Figure 5-2).  Relative to the base case, softwood inventories in 2040 rise by about 5 percent, and
hardwood inventories rise by 7 percent. Overall harvest levels decline, as illustrated in Table 5-2. The
magnitude, as well as the pattern, of changes in harvests reflects the fact that some of the "savings" in
wood that would have been used as pulp for manufacturing paper and paperboard are being used for
the manufacture of other products, especially lumber, after the time lag needed for pulpwood-sized
trees to grow to sawtimber size.
                                        Table 5-2
                Changes in Harvest Levels for the Increased Recycling Scenario,
                         Relative to the Base Case (million cubic feet)
Region
North
South
Rockies
Pacific Coast
Total U.S.
2000
-154
-304
0
-14
-473
Hardwood
2020
-171
-285
0
-14
-470
2040
-123
-172
0
-8
-304
2000
-60
-580
-5
-160
-805
Softwood
2020
-60
11
-48
-187
-284
2040
-40
300
-17
-128
117
 Note: Sum of regions may not equal totals because of rounding.
 Source: Haynes, et al. 1994
                                               37

-------
            Figure 5-2.  U.S. Softwood and Hardwood Inventory Projections for Privately Owned
                       Timberland: Base Case and Increased Recycling Scenario
          I
-Base Case, Softwood
• Base Case, Hardwood
•Increased Recycling,
 Softwood
-Increased Recycling,
 Hardwood
                Source: Haynes, et al. 1994
    ECONOMIC AND TRADE EFFECTS

    Because additional recycling reduces the demand for pulpwood, stumpage prices and revenues
to forest land owners decline over time under the increased recycling scenario relative to the base
case. Reduced harvest for pulpwood increases the future supply of sawtimber, thereby increasing
harvests of sawtimber and decreasing prices (see Table 5-3). Increases in the supply of sawtimber do
not happen immediately, but may lag more than 10 years behind the initial impacts of recycling on
pulpwood markets. Results are felt most strongly in the softwood and hardwood markets in the South,
which produces the bulk of the paper pulp consumed in markets that switch to recycled fiber.

    Similarly to the tree planting scenarios, reduced sawtimber prices in the U.S result in decreased
imports from Canada. Softwood lumber imports from Canada drop significantly throughout the time
period and by over 4,600 million board feet in 2040, or 36 percent of base case imports in that year
(Haynes, et al.  1994).

    5.2 REDUCED NATIONAL FOREST HARVESTS

    The analysis of changing National Forest harvest levels examines the effects of harvest reductions.
These reductions result from elimination of harvest of old growth in the Pacific Northwest, protection of
spotted owl habitat in Washington, Oregon and California,  protection of the red  cockaded woodpecker
in the South, elimination of below cost timber sales, and elimination of harvesting in existing roadless
areas. As illustrated in Table 5-4, the overall level of National Forest harvest for the scenario is about
1.36 billion cubic feet in 2000 and 1.59 billion cubic feet in 2040, about 21 percent below harvest
levels in the base case. Harvest levels projected under the base case, for all public and private lands,
are presented by region in Table 5-5. Most of this analysis relies on the base case and models used
for the  analyses of large scale tree planting presented in Section 4.1, and the CGTM provides
additional results.
                                               38

-------
                                     Table 5-3
                   Stumpage Prices In the Increased Recycling Scenario
                            (dollars per thousand board feet)
2000
Softwood Stumpage Prices
North 48
South 125
Rocky Mountains 99
Pacific Coast 241
2010

50
6
197
274
2020

63
29
190
274
2030

67
70
191
252
2040

75
165
202
236
 Hardwood Sawtimber Stumpage Prices

   United States           461          551
                                     651
                                 764
                                                                            879
 Source: Haynes, et al. 1994
                     Table 5-4
     National Forest Harvest Under Base Case and
       Reduced National Forest Harvest Scenario
                 (billion cubic feet)
    Year
Base Case
Reduced Harvest
    Scenario
1986
2000
2010
2020
2030
2040
2.07
1.54
1.64
1.70
1.74
1.78
2.07
1.05
1.08
1.09
1.11
1.11
 Source: Haynes, et al. 1994
    RESULTS; CARBON, TIMBER INVENTORY, ECONOMIC EFFECTS, AND TRADE

    Timber inventory and carbon impacts for this scenario are modeled using the TAMM/ATLAS
system and the forest carbon models, FCM and FORCARB. The CGTM model provides additional
analysis of the impact of international trade linkages. Under this scenario, declines in timber
inventories result in intensified competition for the available timber and upward pressure on softwood
Stumpage prices. In regions where there are sufficient private timber supplies, however, decreases in
National Forest harvest are offset by increased timber harvests from private timberlands.

    Overall, there is little change in the total timber inventory and, hence, in the total carbon stored on
U.S. timberland. Because of offsetting private harvests, total public and private harvest remain
relatively unchanged under this scenario, as does the carbon stored in the product and landfill pools.
It is important to note (as described in Box 4-1) that this analysis does not account for changes in
                                              39

-------
management behavior in response to price signals, and likely overestimates the economic impacts of
these policies.

                                         Table 5-5
                           Base Case Harvest Projections by Region:
                             Public and Private Timberland Owners
                                      (billion cubic feet)
Region/Owner
North
Public
Private
South
Public
Private
Rockies
Public
Private
Pacific Coast
Public
Private
Total U.S.
Hardwood Harvest
2000 2020 2040

.2
1.8

.1
2.8

—
--

.1
.1
5.1

.3
3.8

.2
5.3

—
-

—
.3
g.g

.4
4.9

.3
5.3

—
--

—
.4
11.3
Softwood Harvest
2000 2020 2040

.1
.6

.4
5.3

.5
.3

1.8
2.1
11.2

.2
1.0

.4
5.6

.6
.5

1.3
2.3
11.8

.2
1.6

.5
7.5

.7
.6

1.3
2.9
15.4
  Note:  Hardwood harvest is reported for the West, i.e., Rockies and Pacific Coast combined.
  Source: Haynes, et al. 1994
    Private softwood inventory (which is illustrated in Figure 5-3) declines by about 1 to 2 percent over
the time period relative to the base case. This change in available timber supply affects the prices of
sawtimber and timber products. Softwood stumpage prices in all regions are significantly affected
through 2040, as a comparison of Tables 5-6 and 4-5 illustrates. Softwood stumpage prices on the
Pacific Coast, for example, are about 12 percent above the base case by the year 2040. Softwood
lumber prices, overall, are about 7 percent higher in 2040 than in the base case (Haynes, et al. 1994).
There are no significant impacts on hardwood timber inventories or hardwood  saw timber stumpage
prices,  illustrating the small role of National Forests in the hardwood sector.

    Because of the lumber price increases, softwood lumber consumption declines slightly, by about
16 percent, and lumber imports from Canada rise by 10 percent by 2040 (Haynes, et al. 1994). The
increase in lumber imports comes progressively after 2010 because domestic  production is reduced
as a consequence of the lower timber inventories and the associated higher prices.

    The CGTM analysis provides slightly different results; higher Pacific Northwest log prices prompt
higher-cost producers to increase their production levels both internationally and  domestically. Thus,
reduced harvests on U.S. public lands are replaced not only by large increases in supply from the
U.S. South, but  also by increased production from Europe and possibly from mature softwood stands
in Siberia. In addition, CGTM projects non-wood product substitution to meet as much as one-third of
the supply reduction (Perez-Garcia 1993).

    The CGTM results have several implications for the carbon and economic  impacts of reducing
 harvest levels in National Forests. The analysis using TAMM/ATLAS  and FCM/FORCARB finds that
 increased harvests on U.S. private timberland replace decreased harvests on  public lands, resulting in
 little net effect on total U.S. inventories or carbon flux from all U.S. timberland. The CGTM analysis
                                              40

-------
                        Figure 5-3. U.S. Private Timber Inventory: Changes Relative
                                   to Base Case for National Forest Harvest Scenario
         1 -r
                                                                           Softwood South
                                                                            r
                                                                          "
                                                                           x
                                                                              Softwood, All Regions
                                                                                \
                         2000
           Source: Haynes, et al. 1994
                                         2010
                                                         2020
                                                                         2030
                                                                                         2040
                                                   Year
finds, however, that a significant portion of the offsetting harvests occurs in countries other than the
U.S. (see Figure 5-4). Thus, carbon storage in U.S. forests could be higher in the reduced harvest
scenario than under the base case.13 The impacts on global carbon, however,  are unclear and
depend on the relative productivity of the lands harvested in the U.S. and internationally, and on the
carbon impacts of substituting more energy-intensive products for forest products.

    The CGTM analysis also shows economic surplus gains of $1.4 billion to timber producers
worldwide. U.S. producers experience a net welfare loss of almost $100 million. Economic losses to
lumber mills in the U.S. West are about $270 million, which are only partially offset by gains to mills in
the South and North of $56 million. Consumer surplus losses in the U.S., which total $971 million, are
evenly distributed among all regions (Perez-Garcia 1993).

    5.3 INCREASED USE OF BIOMASS ENERGY

    This analysis examines the implications of using existing forest resources in the U.S. to increase
the use of bioenergy for electricity generation. This analysis derives the wood energy (biofuel)
   "1 *3
      The CGTM analysis assumes that public harvests decline significantly in Western Canada, thereby reducing
potential timber supplies to the U.S.


                                               41

-------
                                   Table 5-6
          Stumpage Prices In the Reduced National Forest Harvest Scenario
                         (dollars per thousand board feet)
                       2000
2010
                                                2020
                       2030
                                                                        2040
Softwood Stumpage Prices

  North                  53

  South                 208

  Rocky Mountains        159

  Pacific Coast           288
 72

 322

 276

 357
Hardwood Sawtlmber Stumpage Prices

  United States           463         559
98

370

315

405



665
114

331

305

409



784
129

342

329

401



903
Source: Haynes, et al. 1994
                Figure 5-4. Projected Global Production Increases in Response
                          to Habitat Preservation, by Region: 1995
                            38%
                       21%
                                                 20%
                         Q Europe

                         IBJapan

                         E3 Eastern Canada

                         E3 US North

                         0 US South

                         • US West

                         D Product Substitution
                                     12%
                 Note: Distribution reported as a percent of total production increase of
                 33.3 million cubic meters.
                 Source: Perez-Garcia 1993
scenario from the National Energy Strategy (NES) (USDOE 1991). The scenario assumes that wood
energy supply increases from 3.1 Quads in 2000 to 5.4 Quads in 2030, compared to the base case in
which wood energy supply increases from 1.8 Quads to 2.0 Quads over the same period (Colin and
Skog 1990). The scenario accounts for changes both in fuelwood consumption and in the proportion
of fuelwood that comes from the growing stock part of timber inventories (USDOE 1991). Harvest
projections for the base case and scenario are shown in Table 5-7.
                                               42

-------
                                  Table 5-7
                  Biomass Fuelwood Consumption: Assumptions
              for Base Case and Biofuel Scenario, Decadal Intervals
                          Base Case                  Biofuel Scenarios

      Year          Softwood      Hardwood     Softwood      Hardwood
  Biomass Fuelwood Consumption (million cubic feet)
2000
2010
2020
2030
2040
883
1,207
1,312
1,271
1,219
3,377
4,395
4,420
4,113
3,875
1,030
1,650
2,160
3,040
3,200
3,940
6,020
7,200
9,890
10,070
  Fuel from Non-growing Stock Sources (percent)
2000
2010
2020
2030
2040
85
85
86
85
86
75
76
77
77
77
73
82
70
61
67
73
82
70
61
67
 Source: USDOE 1991, reported in Haynes, et al. 1994
    The growing stock fraction is especially important in dealing with fuelwood demand on forest
resources. Of the approximately 737 million acres of forest land in the  U.S., 47 million acres are
reserved from  harvest, and 200 million acres  are urban or low productivity forest land, leaving 490
million acres to support harvests for various products (Haynes, et al. 1994). Harvests for fuelwood
primarily come from urban forest land, low productivity forest land, or from residues left after logging.
There has, thus, been little actual harvest for fuelwood in some regions and, overall,  the majority of
both hardwood and softwood fuelwood comes from non-growing stock, both currently and under the
base case and scenario projections. The percentage of fuelwood derived from growing stock (wood of
merchantable quality) is projected to increase in the NES scenario for  both softwoods and hardwoods.
That increase,  combined with the higher overall harvest level has dramatic effects on forest resources,
especially hardwoods.

    TIMBER INVENTORY EFFECTS AND CARBON IMPACTS

    Increasing the production of energy from biomass has three distinct effects on CO2 emissions:
(1) CO2 is released when the wood is combusted to produce energy; (2) carbon is sequestered
during the growth or regrowth of forests harvested for biofuels; and  (3) carbon emissions are
"avoided" when fossil fuel combustion is displaced by bioenergy. Removing wood for biofuels initially
has the effect of reducing the carbon inventory on the forest base.14 Over time, however, regrowth
on harvested lands replaces the biomass carbon. The real benefit to the atmosphere occurs as wood
displaces fossil fuels, and the CO2 emissions from burning wood are absorbed by regrowing forest.
     Strategies that plant dedicated energy crops will initially increase the carbon inventory prior to harvesting,
unlike strategies that harvest existing forests.

                                              43

-------
              Figure 5-5. Total Carbon Storage on U.S. Public and Private Timberland:
                        Changes Relative to Base Case for Increasing Biomass Energy Scenario
                                         2010
                                                      2020
                                                                    2030
                                                                                  2040
                                                Year
                           2000
               Sources: Lee 1993c; Heath 1993
   The considerably greater harvest levels associated with increased biofuels utilization result (by
2040) in a large reduction in the U.S. timber inventory and, thus, the total carbon pool, relative to the
base case. Estimates of the changes in carbon storage and inventory are shown in Figures 5-5 and
5-6, respectively. The large declines in timber inventory and carbon displayed in these figures result
from several factors. First, the additional biofuel harvest levels under this scenario are large; in 2010,
additional biofuel harvest over the base case is about 37 percent over the base case, and as large as
the total harvest from National Forest lands in 1990. These increases in harvest leave gaps in the
inventory that cannot be filled until the forest regenerates. Because hardwoods, which in general
regenerate more slowly than softwoods, supply much of the harvested timber, regrowth lags
significantly behind harvest, and the inventory shows large declines throughout the time period.

   Second, the volume of harvest rises almost uniformly throughout the time period of the scenario
analysis, relative to the base case. As more volume (and more acreage) is being harvested in each
succeeding decade than is being regenerated, the regeneration cannot keep pace with the size of the
harvest. This result is mirrored in the increasingly steeper downward trajectory of carbon storage
losses in the first four decades of Figure 5-5. As the size of the increase abates towards the end of
the simulation period, regeneration begins to catch up with the harvest, and carbon losses from
inventory level out in Figure 5-5. The depletion of inventory and the carbon pool could be mitigated to
some degree, but  not prevented, by taking more of the harvest volumes from southern softwoods,
which grow and regenerate faster than hardwoods.

    Last, as discussed  in Section 3, the analysis does  not take into account the carbon contained in
timber stands before they reach merchantable size (see Box 3-2). Accounting for biomass in stands in
the first few years  after harvest might therefore reduce the size of the  depletion in inventory. A rough
calculation of carbon in young stands suggests that the results of this scenario analysis are not
changed significantly by such inclusion  (Birdsey 1995).
                                               44

-------
                       Figure 5-6.  U.S. Private Timber Inventory:  Changes Relative to
                                  Base Case for Increasing Biomass Energy Scenario
                                                                               2040
                  Source: Haynes, et al. 1994
    On balance, the results suggest that large increases in the rate of removal of standing inventory
immediately deplete forest inventories and carbon pools, and that depletion effect is maintained or
magnified when the size of the harvest increases over time. The carbon pool is likely to return to a
point of no net loss, compared with the baseline, when harvest volumes reach a steady state,
although the analysis was not carried out far enough to confirm this.

    AVOIDED FOSSIL FUEL EMISSION: NET CARBON IMPACTS

    Although the combined forest inventory and carbon models do not produce estimates of the
impact on CO2 emissions of displacing fossil fuels in the bioenergy scenario, projected harvest levels
can be used to estimate these carbon benefits. The magnitude of avoided emissions depends on a
number of factors, including the energy content of the biomass, the heat rate of the biomass-fired
energy technology, and the carbon intensity of the displaced fossil fuel. Cumulative emissions avoided
by the year 2040 could, however, under a reasonable set of  assumptions, range between 470 and 840
million metric tons of carbon with a mean of about 650 million metric tons (ICF 1995).

    Including avoided emissions does not completely offset the reduced carbon storage due to
increased harvest until  the final decade reported here. Table 5-8 presents annual carbon accumulation
in forest ecosystems at decadal intervals, avoided carbon emissions from  displaced fossil fuel use, as
well as total  net accumulation under this scenario. Savings of fossil fuel emissions do not completely
offset forest  carbon losses because fossil fuels (especially coal)  are more  efficient, or have a higher
energy content per ton of carbon, than wood. Consequently, more tons of wood must be burned to
create the same amount of energy produced by a given tonnage of fossil  fuel.

    This result makes an important  point. Even when considering the  contribution of an offset in the
use of fossil  fuels, harvesting an additional portion of existing forest resources for bioenergy depletes
forest carbon pools, and will not fulfill greenhouse gas mitigation objectives in the short term, and
                                              45

-------
                                   Table 5-8
  Annual Carbon Accumulation on U.S. Public and Private Tlmberland Relative to Base
   Case for Increased Use of Blomass Energy, Decadal Intervals (million metric tons)
                         2000
2010
                                                       2020
                                                                     2030
Forest Carbon
Avoided Fossil
Fuel Emissions
Net Carbon
Flux
-7.3
2.6

-4.7

-17.8
7.5

-10.3

-29.6
13.2

-16.4

-21.1
27.5

6.4

 Note: Forest carbon estimates are derived as an average of FORCARB and FCM results. Avoided fossil
 fuel omissions are an average of results assuming the displacement of coal and gas-fired generation.
 Sources: Birdsey and Heath, 1993; Lee 1993; ICF 1995
possibly in the medium term. Once the forest returns to a steady state where removals and
regeneration balance, the fossil fuel offset would probably be a net contribution to greenhouse gas
abatement. The results suggest that a forest bioenergy strategy that uses wood from plantations
established in advance for growing biofuel stocks would mitigate the adverse impacts on forest
inventories and carbon pools of a significant biomass harvest. Combination scenario 3, below,
addresses that situation. Additional planting and regrowth would further mitigate the impacts of
harvest on carbon storage. Forestation can also mitigate increases in stumpage prices that occur in
the biomass energy scenario, presented in Table 5-9.
                                       Table 5-9
                 Stumpage Prices in the Increased Biomass Energy Scenario
                             (dollars per thousand board feet)
                           2000
2010
                                                    2020
                          2030
                                                                              2040
  Softwood Stumpage Prices
     North                   52           76
     South                   183         278
     Rooky Mountains         106         233
     Pacific Coast            254         332
  Hardwood Sawtimber Stumpage Prices
     United States            465         565
              103
              345
              291
              392

              685
125
341
342
420

827
150
358
339
425

947
  Source: Haynes, et al. 1994
    5.4 EFFECTIVENESS OF COMBINED SCENARIOS
    In reality, climate change mitigation activities are unlikely to be implemented singly or in a policy
 vacuum. The U.S. Climate Change Action Plan, for example, identifies some 44 climate change
                                                46

-------
 mitigation activities for simultaneous implementation (White House 1993). The analyses considered
 thus far in this report — as well as most analyses of climate change mitigation options — examine the
 carbon, timber, and economic impacts of isolated policies. The impacts of policies implemented jointly,
 however, may differ from those suggested by inspecting the results for individual policies, because of
 complementary, synergistic, or competing effects across the policies. To investigate the impacts of
 broader forest sector strategies, this report analyzes several additional scenarios that combine policies
 reported earlier in Sections 4 and 5.

    Analyzing combined scenarios allows an evaluation of the relative success or failure of jointly
 executed programs in achieving the competing goals of maximum carbon sequestration and minimum
 economic disturbance. There are, thus, two key reasons to analyze the impacts of simultaneous
 implementation of forest sector policies. First, combining policies can increase the aggregate carbon
 sequestration that is achieved. Because most of the forest sector policies considered in this report act
 to expand the forest carbon sink, combining policies is likely to have additive impacts on aggregate
 net carbon sequestration. Whether these  impacts are, indeed, additive (or less or more than additive),
 can be a consideration in appropriately devising,  and analyzing the cost-effectiveness of, policy
 strategies to sequester carbon in the forest sector.

    Second, many of the policies, taken in isolation, potentially have adverse economic impacts on at
 least some groups in the forest sector, because of price and other market effects (such as changes in
 trade flows). Combining policies that have opposite impacts, e.g., those that create upward pressure
 on prices by  reducing timber supply and those that create downward pressure on prices by reducing
 timber demand, can  result in less economic disruption than occurs under the individual policy
 scenarios. Again, confirming this result and understanding the net impacts of these offsetting effects
 solidifies the analytical underpinnings of the analysis of the economic impacts of forest sector carbon
 sequestration actions.
                                        Table 5-10
                                   Combination Scenarios
      Scenario
Reduced National
 Forest Harvest
                                         Increased       Biofuels         Low Enrollment
                                         Recycling                     $220 million funding
 Combination 1

 Combination 2

 Combination 3
    Combining scenarios, thus, presents an opportunity to evaluate the carbon, timber, and economic
consequences of a coordinated forest sector strategy to sequester carbon. The three combination
scenarios analyzed are described in Table 5-10. Strategy 1  combines reduced National Forest harvest
with increased paper recycling. These two policies have positive or no (and, hence, complementary)
impacts on carbon accumulation  and opposite  (and, hence, offsetting) impacts on prices and
economic welfare. Strategy 2 combines the first strategy with the increased production of biomass
energy; this addition would tend to reduce the carbon stored on timberland (relative to Strategy 1)
and to aggravate the upward pressure on prices created by reduced harvest of National Forests. The
third strategy adds large-scale tree planting to the second strategy.15 The addition of this scenario
would tend to offset the declines  in timber inventory and carbon storage associated with the biomass
energy scenario, and provide offsetting  downward pressure on softwood stumpage prices.
   15
     ' The afforestation program added to the Combination 3 analysis is the low enrollment distribution (Moulton
and Richards 1990) with an annual funding level of $220 million for 10 years.
                                              47

-------
    IMPACTS ON TIMBER INVENTORY AND CARBON ACCUMULATION

    Figure 5-7 displays the impacts of the combined scenarios on total carbon storage, using the
results from the FCM carbon model. Figures 5-8 and 5-9 display the underlying changes in U.S.
private softwood and hardwood timber inventories over the period of the analysis. Because there is
little change in net carbon accumulated over the base case in the reduced harvest scenario, the
carbon impacts of the increased paper recycling scenario, reported earlier in Section 5, are similar to
those of combination 1 (illustrated in Figure 5-7). As would be expected, the incremental carbon
stored, relative to the base case, is  lower for combination 2 than for combination 1,  because of the
addition of the increased biomass energy scenario, which reduces the timber inventory substantially.
Adding an aggressive tree planting  program (the third combination) further increases the carbon
benefits of forest sector policies.  Table 5-11, which presents net annual carbon accumulation in
the forests at decadal intervals for the FCM model, displays analogous results.
                      Figure 5-7.  Total Carbon Storage on U.S. Public and Private Timberland:
                                 Changes Relative to Base Case for Combined Scenarios
                                                                        Combination 1
                                                                    2030
                                                                                2040
                   Source: Lee 1993c
   16 Using the high, rather than the low, enrollment assumptions for the $220 million afforestation program in
combination 3 would not change the overall results dramatically, although the net accumulation of carbon would
exceed that in Figure 5-7.

                                               48

-------
    Figure 5-8.  U.S. Private Softwood Timber Inventory: Changes
                 Relative to Base Case for Combined Scenarios
 20-r
                                                          Combination 3
                                            Combination 2
               2000
                           2010          2020
                                 Year
                                                    2030
                                                                 2040
   Source: Haynes, et al. 1994
       Figure 5-9.  U.S. Private Hardwood Timber Inventory:  Changes
                    Relative to Base Case for Combined Scenarios
 20--

 15- •

 10- •

 5--

 0- -

 -5--

-10 • •

-15--

-20--
-25 J-
  1990
                  Combination 1
Combination 2
                 Combination 3
               2000
                            2010          2020
                                  Year
                                                      2030
                                                                   2040
    Source: Haynes, et a!. 1994
                                  49

-------
                                          Table 5-11
       Annual Carbon Accumulation on U.S. Public and Private Timberland Relative to Base Case
                   for Combined Scenarios, Decadal Intervals (million metric tons)

Combination 1
Combination 2
Forest Carbon
Avoided Fossil
Fuel Emissions
2000
10.6

4.2
2.6
2010
13.0

-3.8
7.5
2020
8.8

-23.8
13.2
2030
3.7

-24.2
27.5
  Net Carbon Flux

 Combination 3

  Forest Carbon

  Avoided Fossil
  Fuel Emissions

  Net Carbon Flux
6.8


10.1

2.6


12.7
3.7



11.0

7.5


18.5
-10.6



 -6.6

 13.2


 6.6
 3.3



-12.7

 27.5


 14.8
 Noto: Forest carbon estimates are derived from the FCM model. Avoided fossil fuel emissions are an average of results
 assuming the displacement of coal and gas-fired generation.
 Sources: Lee 1993c; IGF 1995
    ECONOMIC IMPACTS

    In Table 5-12, stumpage prices over the time frame of the analysis are provided for each of the
combined scenarios. In combination 1, stumpage prices for softwood, are higher than for the recycled
fiber scenario and lower than the harvest reduction scenario. Prices do not, however, return back to
base case levels, demonstrating (in this case, at these levels of activity) that increased use of recycled
fiber utilization has more of an effect on prices than do the public harvest reductions. Similar trends
are evident for these scenarios in consumer and producer surplus, and stumpage producer
(landowner) surplus (Haynes, et al. 1994).

    Adding a large biomass energy program based on existing wood resources from the forest sector
increases prices over the first combination, as the large draw on the forest resource for biofuel feed
stocks depletes timber supplies. The magnitude of the price change varies by region, with those
regions supplying the most additional biomass for energy experiencing the largest price increases.
Stumpage suppliers gain somewhat in economic welfare,  but other economic players (consumers and
product producers) show almost no change in welfare compared to the base case (Haynes, et al.
1994).

    Finally, the third combination scenario adds an aggressive tree planting program. Economic
impacts vary, with the most severe effects experienced in South for the latter decades of the analysis,
when the large increase in timber inventories  from federally subsidized tree planting undermines the
stumpage market and drives prices down. This analysis suggests that a  bioenergy program that
draws extensively on  Southern timber markets could raise demand enough to partially offset the
negative impacts of a large tree planting program. The addition of a tree planting program mitigates
the price increases stimulated by the bioenergy program  in the second combination, and prices in
some regions and years falls below those in the base case. Here again,  the presence of other
activities in the combination reduces the adverse economic impacts of a tree planting program
implemented by itself.
                                               50

-------
                                          Table 5-12
                           Stumpage Prices in the Combination Scenarios
                                 (dollars per thousand board feet)
2000
Combination 1
Softwood Stumpage Prices
North
South
Rocky Mountains
Pacific Coast
Hardwood Sawtimber Stumpage
United States
Combination 2
Softwood Stumpage Prices
North
South
Rocky Mountains
Pacific Coast
Hardwood Sawtimber Stumpage
United States
Combination 3
Softwood Stumpage Prices
North
South
Rocky Mountains
Pacific Coast
Hardwood Sawtimber Stumpage
United States


52
161
155
280
Prices
461


53
161
150
282
Prices
462


53
161
161
286
Prices
462
Note: "*" indicates prices outside the range of model
2010


56
67
224
314

551


71
102
241
326

557


71
78
220
327

557
results.
2020


73
111
238
317

651


100
134
261
339

671


100 .
30
245
324

668

2030


80
168
264
305

764


125
202
329
341

813


125
*
273
299

810

2040


94
253
232
287

879


157
261
333
337

935


157
*
232
280

931

 Source: Haynes, et al. 1994
6.  MODIFIED AGRICULTURAL PRACTICES

    Agricultural soils store an estimated 1.5 trillion metric tons of carbon, which is about twice the
amount held in the atmosphere (Post, et al. 1990). Modifying current agricultural practices to increase
or maintain this carbon sink may have a significant effect on the flux of carbon between agricultural
soils and the atmosphere. Possible alternative management strategies include modifying tillage
practices, using winter cover crops, changing crop rotations, and altering the extent or composition of
                                              51

-------
fertilizer use. Specifically, this section analyzes two types of scenarios: (1) increased use of
conservation tillage practices; and (2) increased planting of winter cover crops. Conservation tillage
can reduce the loss of soil organic carbon by decreasing the soil disturbance of conventional tillage;
using winter cover crops can increase the stored carbon in agricultural soils by expanding biomass
production on existing farm lands.

   These practices can also affect the release of soil N2O. Although the magnitude of N2O emissions
is much smaller than that of CO2, because the global warming potential of N2O is 270 times that of
CO2 the potential impacts on global warming may be significant. While preliminary results indicate that
these policies may also reduce N2O emissions  (Cialella and Li 1993), because of the complex
interactions between carbon and nitrogen emissions from agricultural soils no quantitative results are
presented.

   The analysis of agricultural management practices was conducted by integrating the RAMS
economic model with the CENTURY soil carbon model, with the use of GIS analysis to combine
results and incorporate meteorologic and soils databases. As described in Section 3, the RAMS  model
is a short-term, profit maximizing model,  which is used in this analysis to determine detailed
production activities (e.g., crop rotations, tillage, and irrigation) under the constraints of the scenarios
chosen for this report. The production activities determined in the RAMS analysis for each area of the
study region are used by the CENTURY model to estimate the associated soil carbon impacts with a
high level of regional detail. The DNDC (Denitrification and Decomposition) model was used to validate
the results of the CENTURY model, and  to assess potential impacts on N2O emissions.

   The RAMS model focuses on the Midwestern and Central U.S. (see Figure 6-1), which comprises
about 216 million acres, or 60 to 70 percent of the agricultural cropland in the contiguous U.S.
(Donigian, et al. 1994). Because factors that influence soil carbon and economic impacts, such as soil
type, season length,  and management practices (e.g., crop rotations), vary significantly from region to
region, the results presented below cannot be directly generalized to all regions in the U.S.
Nevertheless,  given the magnitude of the RAMS study region, they indicate that modifying agricultural
management practices can significantly  increase the accumulation of soil carbon.

    6.1  MODIFIED TILLAGE PRACTICES

    Modifying tillage  practices can have  significant impacts on the sequestration of carbon in
agricultural soils, soil erosion, and economic welfare.  In this analysis, four types of tillage practices are
modelled: conventional spring and fall tillage, reduced tillage, and no-tillage. These four practices
differ in important physical aspects, such as the frequency and depth of plowing, as well as other
tillage operations, resulting in different transfers of carbon and nitrogen between various above and
below ground pools. Depending on a range of regional factors captured in the models, such as  crop
rotation, soil type, climate, and management practices, the reduced disturbance of the soil using
conservation tillage practices (i.e., reduced tillage and no-tillage) may increase the retention of soil
organic carbon relative to conventional tillage.

    The impacts of these tillage practices are assessed using four scenarios that represent
progressively  greater acreages of farmland managed with conservation tillage practices. Figure 6-2
shows the degree to which each tillage practice is used in each scenario. These scenarios were
developed by targeting the application of conservation tillage practices on agricultural lands with the
highest erodibility indexes (El). By including land with progressively lower El ratings (i.e., implementing
conservation tillage on progressively less erodible land),  more land is managed with these practices.
The  low, medium, and high conservation scenarios shown in Figure 6-2 have  minimum El limits of 8, 5,
and  2, resulting in the inclusion of 18, 27, and 53 percent of U.S. agricultural land, respectively. The
targeting of highly erodible lands in these policy scenarios has the additional  benefit of reducing soil
erosion on agricultural lands.
                                              52

-------
                                  Figure 6-1.  RAMS Study Region
        Source: Donigian, et al. 1994
   The base case (status quo) scenario, as defined by the RAMS model, is based on data for the
1990 growing season. The use of conservation tillage may, however, already exceed levels
comparable to the medium conservation scenario. If the actual "no-policy" base case has a higher use
of conservation tillage than the base case used in the current analysis, then aggressive policies  (i.e.,
high conservation) will in fact be easier and less costly to implement, although carbon sequestration
benefits will also be reduced. Nevertheless, the results presented in the next section indicate the
potential importance of promoting conservation tillage.

   CARBON IMPACTS: SOIL ORGANIC CARBON™

   In general, the modelssshow a net accumulation of carbon in agricultural soils within the RAMS
study region, resulting in an  increase in soil organic carbon of 26 to 52 percent by 2030, depending
on tillage practices,  crop yield increases,  and other assumptions. The low conservation scenario
results in roughly a  1 percent increase in soil carbon levels relative to the base case by 2030, while
the medium and high scenarios result in 2 and 6 percent increases, respectively. These increases
correspond to average annual carbon accumulations above base case levels of 0.25, 0.5, and 2.5
million metric tons for the low, medium, and high conservation scenarios, respectively  (see Table 6-1).
   17 Soil organic carbon refers to carbon in the first 20 centimeters of the soil profile, including roots and surface
residues.

                                              53

-------
                               Figure  6-2.  Tillage Distributions for Alternative Scenarios
                                Status Quo/Base Case
 Low Conservation
                               Medium Conservation
High Conservation
                 Source: Donigian, et al. 1994
                                              Table 6-1
                Soil Organic Carbon Accumulation for the Conservation Tillage Scenarios
                            in the RAMS Study Region  (million metric tons)
Scenarios
Base Case
Low Conservation
Medium Conservation
High Conservation
Carbon
Accumulation
1990-2030
1,800
1,810
1,820
1,900
Percent
Increase from
1990 Value
49%
49%
50%
52%
Percent Increase
above Base
Case
--
1%
2%
6%
Average Annual
Carbon
Accumulation Above
Base Case
--
0.25
0.50
2.50
 Noto: These results assume an annual crop yield increase of 1.5%,
 Source: Donigian, et al. 1994
    The potential of conservation tillage practices to increase carbon sequestration is greatly affected
by factors such as crop yield, crop rotation, and regional variations in soil type, geography, and
season length. As a result of these variations, the aggregate results presented here cannot be
generalized to specific land areas. The impacts of any policies to encourage conservation tillage will
                                                 54

-------
depend greatly on which agricultural lands, and associated production activities, are targeted. Box 6-1
provides an illustration of the regional variation of crop rotations and tillage practices, and their
importance for soil carbon results. The remainder of this section discusses how the analysis
addressed each of these factors, and their impact on soil carbon results.
                                             Box 6-1
                          The Effects of Crop Rotation, Tillage practice, and
                     Climate on Soil Carbon for Two Neighboring Climate Divisions
         As an example of how sail carbon varies wtth crop rotation, tillage practice, and Ottmate Division,
     consider the example of two neighboring Climate Divisions. CD 312 is in the eastern-most part of the
     stuoy region on the border of Ohio and West Virginia. This CD attains its highesl levels of soft carbon
     for a rotation of corn Sitage-corn silage-Soybean, for which, by the year 2030, soil carbon levels reach
     10fQ96 gC/m2 for Reduced Tilt and tO,400 gC/ra2 for Mo-Till. When this rotation is replaced by Com-
     Corn*$0ybean* However, soil carbon Is only 6,925 gC/m2 under Deduced Tl and 7,221 gC/m2 under
     NO-Til).                 ,         ,            	

         Considering the exact same two crop rotations and tillage practices-in the neighboring Climate
     Division 313 (same Production Area, different climates, on the Wesi Virginia-Kentucky border) produces
     a significantly different story*  For the Corn Silage-Corn Silage-Soybean rotation, by the year 2030, soil
     carbon levels reach 8,413 gG/m2 for Reduced Till-and 8,526 gC/m2 for No-TiB. When this rotation is
     replaced by eorn-com-soybean,. however* sol carbon is only 5,415 gc/mF under Reduced Till and
     5,758 gG/m2 under No-Till.

         It should be noted, however, that even the initial soil carbon levels were somewhat different for the
     two CDs, in 1880, soil carbon was estimated at 3,543 gC/m2 irt OD.,312, white in OBu3i3f SOG.was  	
     only 3,105 o,G/nr<  Thus, although soil carbon increases in CO 312 were indeed higher in percentage
     terms than in CD 312y they were not quite as high as the'absolute projections would otherwise indicate.

     Source: Ooniglar), et aft. 1994       _,             ,,              ,              ^
    Crop Yield. Crop yield can have several important effects on the soil carbon impacts of
conservation tillage policies. First, potential short term crop yield decreases following the introduction
of conservation tillage may deter individual producers from changing tillage practices, and also may
influence the choice of crop rotations. Second, assumptions about long term annual crop yield
increases have a large impact on total carbon accumulation.

    Although implementing  conservation tillage practices may not have a significant long term impact
on crop yield, lower yield may be expected during the initial adoption period as producers learn to use
the new tillage systems (Cruse 1992). The magnitude of such losses will affect the degree to which
conservation tillage practices are  adopted. With small or negligible economic losses, gains realized
from reduced soil erosion may be sufficient to make conservation policies widely appealing, while
large initial losses may hinder implementation.

    Two alternative assumptions are evaluated in order to assess the impact of short term crop yield
decreases. The first assumption keeps crop yields roughly the same for both conventional and
conservation tillage practices; the second assumes 5  and 10 percent yield reductions for reduced till
and no-till practices, respectively. As shown in Table 6-2, these assumptions affect the acreage of
farmland on which conservation tillage practices are employed, and thus reduce carbon accumulation
within the study region. Crop yield decreases also affect the type of crop rotations used, which,  as
described below, have a  large impact on carbon accumulation. Table 6-3 illustrates the variation in
major crop rotations under different conservation targets and short term yield assumptions.
                                                55

-------
                                           Table 6-2
          Impact of Short Term Yield Assumptions on the Introduction of Conservation Tillage
                        (percent change in acreage of each tillage practice)
                                             Conservation Tillage Targets

                                 Low Conservation                    High Conservation

                         Reduced Yield    No Yield Change    Reduced Yield     No Yield Change
Tillage Practice
 Conventional Tillage/
 Fall Plow
                          -7%
  -2%
-46%
-44%
Conventional
Tillage/Spring Plow
Reduced Tillage
No-Tillage
-4%
9%
48%
-7%
7%
69%
-32%
56%
401%
-33%
8%
905%
Source: Bouzaher, et al. 1993
                                            Table 6-3
                   Impact of Short Term Yield Assumptions on Major Crop Rotations
                          (percent change in acreage of each crop rotation)
      Crop Rotation
 Key: CRN — Corn for grain
     HLH = Legume hay
     NLH = Non-Legume hay
 Source: Bouzaher, et al. 1993
                        Base Case
         Conservation Tillage Targets

Low Conservation            High Conservation

Continuous CRN
CRN CRN SOY
CRN SOY WWT
Continuous HLH
Continuous NLH

8.36
7.11
23.87
8.43
8.34
Reduced
Yield
8.37
7.22
23.92
8.43
8.39
No Yield
Change
8.42
6.03
25.03
8.34
7.94
Reduced
Yield
7.96
7.13
24.23
7.63
6.96
No Yield
Change
8.71
6.39
24.82
8.45
6.65
                          SOY = Soybeans
                          WWT = Winter wheat
    The second analysis examines the significance of annual increases in crop yield during the
simulation period. Higher crop yields will increase soil carbon accumulation because, for a given set of
production activities and conditions, more biomass is produced and subsequently returned to the soil.
A 1.5 percent annual increase in crop yield, based on historic data and projected future yields, was
used for the results presented in this section. For comparison, analyses were also conducted
assuming percentage increases of 1.0 and 0.5. Although total carbon accumulated on agricultural land
rises with the assumed yield level, the absolute amount of carbon accumulated in the scenario over
the base case is relatively insensitive to the yield level (Donigian,  et al. 1994).
                                               56

-------
    Crop Rotation. Carbon sequestration depends on the specific crop rotation that is in production,
in large part because crop rotations return different quantities of biomass to the soil. For example,
crop rotations involving spring wheat sequestered 30 to 60 percent less carbon than continuous corn
planting, and crop rotations involving hay sequestered 60 percent less carbon than spring wheat
(Donigian, et al. 1994; Cialella and Li 1993). Thus, the targeting of specific production regimes, and
effects that alter crop rotations (i.e., short term yield decreases), can influence the soil carbon impacts
of conservation tillage policies.

    Regional Variation. Variation in regional characteristics, such as soil type, geography, and season
length, also affects carbon accumulation in agricultural soils. To analyze this effect, the study region
was divided into 80 climate divisions  (CDs). Table 6-4 provides an illustration of the regional
differences in soil organic carbon  impacts by showing results for five different CDs with the same crop
rotation (corn - soybeans). Even for the two neighboring CDs (CD272 and CD321), percent increases
in soil carbon levels differ by  10 to 20 percent.

                                    Table 6-4
                    Regional Variation In Soil Carbon Accumulation
                     (percent increase in soil carbon, 1990-2030)
  Tillage Practice
  CD272
(NW Ohio)
 CD321
(W Ohio)
   CD392
(S Minnesota)
CD413
(Iowa)
   CD603
(W Missouri)
 Conventional Till/
  Spring Plow
   45%
  53%
    37%
 20%
    64%
Reduced Till
No-Till
50%
61%
58%
68%
40%
52%
23%
36%
69%
78%
 Note: Most CDs cross one or more state boundaries.  The primary state is listed for each CD.
 Source: Donigian, et al. 1994
    AGRICULTURAL IMPACTS: SOIL EROSION AND NET RETURN PER ACRE

    Soil erosion rates decrease under the more aggressive conservation tillage scenarios. For the low
conservation case, soil erosion rates are reduced by approximately 3 percent from base case levels.
In the high conservation case, erosion drops by 23 to 34 percent from erosion  rates in the base case,
depending on the short term crop yield  decreases and other assumptions (Bouzaher, et al. 1993).

    Economic impacts in the agricultural sector depend primarily on assumptions about crop yield
changes. Without yield decreases, net returns increase relative to the base case by $0.67 per acre
(0.8 percent) under the low conservation scenario, and by $4.73 per acre (5.4 percent)  under the high
conservation scenario. If lower yields are assumed with conservation tillage practices, then net returns
decrease by $3.07 per acre (3.4 percent) under the low conservation scenario, and by $2.93 per acre
(3.3 percent) under the high conservation scenario.  These results reflect the change in  both crop
revenues (which decline if yields fall) and production costs (which are lower with conservation tillage).
If yields are assumed to remain constant, then revenues also remain constant and the reduced
machinery costs associated with using conservation tillage practices result in increased net returns. If
yields are assumed to decrease, then the lower revenues associated with decreased yields are not
fully offset  by reduced machinery costs. In either case (i.e., with or without yield adjustments), net
returns per acre are greater under the high conservation scenario than under the low conservation
scenario (Bouzaher et al. 1993).
                                               57

-------
    6.2 WINTER COVER CROPS

    Winter cover crops can increase soil carbon, but are only appropriate for use on a fraction of U.S.
agricultural land due to restrictive crop rotations, growing season lengths, and climates.  Based on
these factors, approximately 5 to 10 percent of the land in the RAMs mode! was targeted as being
suitable to establish winter cover crops. The desired sequence of crop rotation for winter cover crops
was small grains or silage followed by crops not seeded in the fall. Crops seeded in the fall (e.g.,
winter wheat and  leguminous and non-leguminous hay) already provide for winter cover, while the
growing period for winter cover crops was greater following crops with early harvest dates.

    Figure 6-3 provides a breakdown of how crop rotations would change if winter cover crops were
introduced on suitable land. In addition to timing issues (crop rotation, and harvest and  planting
dates), the availability of water is a limiting factor. For example, some planting areas in the West do
not have enough water to support a winter cover crop, whereas the longer growing  seasons in the
South allow  substantial growth.
     Figure 6-3. Major Rotations: Base Case and Targeted Levels of Winter Cover Crops
                                 (percentage of total acres)
             Base Case
  Winter Cover Crops
                              Ccrtnuccs CRN
         Rye Cover
CRN CRN SOY   4%      Continuous CRN
  7%
                                        CBNSOV   Other Rotations
                                         35*      35%
                                                                                   CRN SOY WWT
                                                                                      0%
                                                                                 Continuous HLH
                                                                                    7%
                                                                 ___^___     SMF SWT
                                                               Continuous NLH VMch c™r 5%
       Source: Bouzaher, etal. 1993
    Of the cover crops planted, 55 percent by acreage were rye and 45 percent were hairy vetch,
which is often used to fix nitrogen. In the scenarios evaluated, rye cover was used with corn-soy-winter
wheat, corn silage-soy-legume hay, and sorghum-winter wheat rotations. Hairy vetch cover was
planted with the corn-soybean-winter wheat rotation.

    RESULTS: CARBON, ECONOMIC, AND AGRICULTURAL IMPACTS

    Where appropriate, policies to promote cover crops have the potential to increase soil carbon
levels significantly. Although only a small fraction of the agricultural land would be included in such a
                                               58

-------
scenario, there is a net soil carbon accumulation above the base case of roughly 140 million metric
tons of carbon into soil by 2030, or an average annual carbon accumulation of 3.5 million metric tons
(see Table 6-5).


                                           Table 6-5
                 Soil Organic Carbon Accumulation for the Winter Cover Crop Scenario
                                       (million metric tons)
Scenarios
Status Quo
Winter Cover Crops
Carbon Accumulation
1,800
1,940
Percent
Increase
from 1990
Value
49%
53%
Percent
Increase above
Base Case
--
8%
Average Annual
Carbon Accumulation
above Base Case
--
3.5
  Note: These results assume an annual crop yield increase of 1.5%.
  Source: Donigian, et al. 1994
    Because carbon sequestration is sensitive to the rate at which crop yield is assumed to increase
over time, alternative assumptions of annual increases in crop yield of 0.5, 1.0, and 1.5 percent were
examined. As with the conservation tillage analysis, lower crop yield increases result in similar soil
carbon accumulation relative to the base case, despite lower absolute soil carbon levels by the end of
the study period (Donigian, et al. 1994).

    Average net returns are reduced by 1.2 percent,  or $1.05 per acre.18 Soil erosion is also
reduced by using winter cover crops in the model, with average soil erosion rates decreasing from
4.51 tons per acre to 4.39 tons per acre, a decline of almost 3  percent (Bouzaher, et al. 1993).


7. CONCLUSIONS: DIRECTIONS FOR FUTURE RESEARCH

    The results of the preceding sections suggest that a variety of strategies, in both the forest and
the agriculture sectors, could reduce emissions of greenhouse gases or increase the carbon stored in
biomass and soil. Such policies could also have significant impacts (in some regions or nationally) on
product prices, forest and agricultural production, consumer and producer welfare, and trade. Thus,
research at the U.S. EPA and its collaborators in the U.S. Forest Service,  at universities, and  at private
research organizations, continues to address outstanding questions.

    For the tree planting and other forest policy scenarios, which (with the exception of biofuels)
sequester carbon primarily by enlarging the acreage devoted to trees over time, both economic and
carbon impacts depend critically on how private forest land owners respond to changing inventories
and prices  over time. Economic impacts also depend oh interactions between the forest and
agriculture  sectors, which influence both the cost of  land available for tree planting (and, hence, the
 cost of government-subsidized planting) and also affect consumer  and producer welfare in agricultural
 and forest markets.  Research continues to address these questions, and to investigate issues related
to other aspects of the forest-climate change link, such as the impacts of alternative management
 strategies on public and private lands, improving the data on all portions of the forest sector carbon
 budget, and continuing to evaluate the effects of climate change on forests.
    18 This decrease includes the fact that cost is partially offset by using hairy vetch as a winter crop, since it fixes
 nitrogen, thus resulting in nitrogen savings.

                                               59

-------
    The analyses of policy scenarios that change agricultural practices have focused on a few of the
agricultural areas located in the central United States. Estimating the implications for national GHG
emissions of changing agricultural practices requires understanding how these results can be
generalized to all regions, which may require additional data on selected crops (such as soybeans),
regions, and tillage practices. Other areas requiring further investigation include: improved estimates
of impacts of policies on crop yields; impacts of climate change on carbon sequestration; physical
impacts on the carbon and nitrogen balances of animal waste applications; a more thorough
investigation of the impact of No-Till practices on soil carbon levels; and better tracking of erosional
soil organic carbon losses.
                                              60

-------
                                       REFERENCES
Adams, D.M. and R.W. Haynes. 1980. "The 1980 Softwood Timber Assessment Market Model:
       Structure, Projections, and Policy Simulations." Forest Science. Monograph 22.

Adams, R.M., D.M. Adams, J.M. Callaway, C. Chang, and B.A. McCarl. 1993. "Sequestering Carbon on
       Agricultural Land: A Preliminary Analysis of Social Costs and Impacts on Timber Markets."
       Contemporary Policy Issues 11 (1):76-87.

Birdsey, R.A. 1995. Personal communication to Steven Winnett, U.S. Environmental Protection Agency.

Birdsey, R.A. 1992a. Carbon Storage and Accumulation in United States Forest Ecosystems. General
       Technical Report No. WO-59. Washington, DC: U.S. Department of Agriculture, Forest Service.

Birdsey, R.A. 1992b. "Changes in Forest Carbon Storage from Increasing Forest Area and Timber
       Growth," in Sampson, R.N. and D. Hair, eds., Forests and Global Change, Vol.1. Washington,
       DC: American Forests.

Birdsey, R.A. and L.S. Heath. 1993. Carbon Sequestration Impacts of Alternative Forestry Scenarios.
       Report prepared for the U.S. Environmental Protection Agency. Radnor, PA: USDA Forest
       Service, April.

Bouzaher, A., D.J. Holtkamp, R. Reese, and J. Shogren.  1993. "Economic and Resource Impacts of
       Policies to Increase Organic Carbon in Agricultural Soils." Report prepared for the U.S.
       Environmental Protection Agency. Ames, IA: Center for Agricultural and Rural Development at
       Iowa State University, May.

Callaway, J.M. and B.F. McCarl. 1994.  "Planting Trees  on Agricultural Lands: The Costs and Economic
       Impacts of CRP Reversion." Report prepared for the U.S. Environmental Protection Agency.
       Boulder, CO: RCG/Hagler Bailly, December 14.

Cardellichio, P.A., Y.C. Youn, D. Adams, R.W. Joo,  and J. Chmelik. 1989. "A Preliminary Analysis of
       Timber and Timber Products Production, Consumption, Trade and Prices in the Pacific Rim
       Until 2000." Working Paper 22. Seattle, WA: Center for International Trade in Forest Products
       at the University of Washington.

Chang, C. and B.F. McCarl. 1991. "Scope of ASM: The U.S. Agricultural Sector Model." College
       Station, TX: Texas Agricultural Experiment Station.

Cialella, A. and C. Li. 1994. "Carbon Sequestration and N2O Emissions from Soils: A Model Simulation
       Study for Seven Agricultural Sites in the Central U.S. Using the DNDC Model." Draft report
       prepared for the U.S. Environmental Protection Agency. Washington, DC: The Bruce
       Company,  October.

Colin, H. and K. Skog. 1990. "Current  and Projected Wood Energy Consumption in the United States,"
       in Energy from Biomass and Wastes: Proceedings of the Institute of Gas Technology
       Conference 23:229-260. Chicago, IL: Institute of Gas Technology.

Cruse, R. 1992. Personal  Communication. Department of Agronomy, Iowa State University.
                                             61

-------
Donigian, A.S., T. Barnweli, R.B. Jackson, A.S. Patwardhan, K.B. Weinreich, A. L Rowell, R.V.
       Chinnaswamy and C.V. Cole. 1994. Assessment of Alternative Management Practices and
       Policies Affecting Soil Carbon in Agroecosystems of the Central United States. Report No.
       EPA/600/12-94/067. Washington, DC: U.S. Environmental Protection Agency, Office of
       Research and Development, April.

Haynes, R.W. and D.M. Adams. 1985. "Simulations of the Effects of Alternative Assumptions on
       Demand-Supply Determinants on the Timber Situation in the United States." Washington, DC:
       U.S. Department of Agriculture, Forest Service, Forest Resources Economics Research.

Haynes, R.W., R.J. Alig, and E. Moore. 1994. Alternative Simulations of Forestry Scenarios Involving
       Carbon Sequestration Options: Investigation of Impacts on Regional and National Timber
       Markets. General Technical Report No. PNW-GTR-355. Portland, OR: U.S. Department of
       Agriculture, Forest Service, Pacific Northwest Research Station, August.

Heath, LS. 1993. Personal Communication to Steven Winnett, U.S.  Environmental Protection Agency.
       November 5.

ICF Incorporated. 1995. "Avoided Emissions from Displaced Fossil Fuel Use in the Increased
       Bioenergy Scenario." Memorandum prepared for the U.S. Environmental Protection Agency,
       Washington, DC, April.

Ince, P.J. 1994. Recycling  and Long-Range Timber Outlook. General Technical Report No. RM-242.
       Fort Collins, CO: USDA Forest Service, Rocky Mountain Forest and Range Experiment Station,
       February.

Intergovernmental Panel on Climate Change. 1994. Radiative Forcing of Climate Change:  The 1994
       Report of the Scientific Assessment Working Group of the IPCC: Summary for Policy Makers.
       Geneva: IPCC Secretariat.

Intergovernmental Panel on Climate Change. 1993. Climate Change 1992: The Supplementary Report
       to the IPCC Impacts Assessment. Edited by Tegart, W.J. and G.W. Sheldon. Canberra:
       Australian Government Publishing Service.

Intergovernmental Panel on Climate Change. 1992. Climate Change 1992: The Supplementary Report
       to the IPCC Scientific Assessment. Edited by Houghton, J.T., B.A. Callender and S.K. Varney.
       Cambridge: The Press Syndicate of the University of Cambridge.

Lee, J.J. 1993a.  Personal Communication to Steven Winnett, U.S. Environmental Protection Agency,
       June 23.

Lee, J.J, 1993b.  Personal Communication to Steven Winnett, U.S. Environmental Protection Agency,
       July 6.

Lee, J.J. 1993c.  Personal Communication to Steven Winnett, U.S. Environmental Protection Agency,
       November 3.

Mills, J.R. and J.C. Kincaid. 1992. The Aggregate Timber/and Assessment System —ATLAS: A
       Comprehensive Timber Projection Model.  General Technical Report No. PNW-GTR-281.
       Portland, OR: U.S. Department of Agriculture, Forest Service, Pacific Northwest Research
       Station.

Moulton, R.J. and K.R. Richards. 1990. Costs of Sequestering Carbon through Tree Planting and
       Forest Management in the United States. General Technical Report WO-58. Washington, DC:
       U.S. Department of Agriculture, Forest Service.
                                            62

-------
Parks, P.J. and I.W. Hardie. 1992. "Least-cost Forest Carbon Reserves: Cost-effective Subsidies to
       Convert Marginal Agricultural Lands to Forests." Report to the American Forestry Association.
       Washington, DC.

Perez-Garcia, J.M. 1994. "An Analysis of Proposed Domestic Climate Warming Mitigation Program
       Impacts on International Forest Products Markets." Working Paper 50. Seattle, WA: Center for
       International Trade in Forest Products at the University of Washington, July.

Perez-Garcia, J.M. 1993. "Global Forestry Impacts of Reducing Softwood Supplies from North
       America." Working Paper 43. Seattle,  WA: Center for International Trade in Forest Products at
       the University of Washington, July.

Post, W.M., T.H. Peng, W.R. Emanuel, A.W. King, V.H. Dale, and D.L De Angelis. 1990. "The Global
       Carbon Cycle." American Scientist 78: 310-326.

Reese, R., A. Bouzaher, and J. Shogren. 1993. "Long Term Economic Consequences of Alternative
       Carbon Reducing Conservation and Wetlands Reserve Programs: A BLS Analysis." Report
       prepared for the U.S. Environmental Protection Agency. Ames, IA: Center for Agricultural and
       Rural Development at Iowa State University, May.

Richards, K. 1992. "Derivation of Carbon Yield Figures for Forestry Sequestration Analysis." Draft Paper
       prepared for the U.S. Department of Energy, Office of Economic Analysis. Washington, DC,
       January 8.

Row, C. and R.B. Phelps.  1990. "Tracing the Flow of Carbon Through the U.S. Forest Products Sector,"
       in Proceedings of the 19th World Congress of the International Union of Forest Research
       Organizations. Montreal.

Schneider,  S.H. 1989. Global Warming: Are We Entering the Greenhouse Century? San Francisco:
       Sierra Club.

Trexler, M.  1991. Minding the Carbon Store: Weighing U.S. Forestry Strategies to Slow Global Warming.
       World Resources Institute, Washington, DC.

Turner, D.P., J.J. Lee, G.J. Koerper, and J.R.  Barker. 1993. The Forest Sector Carbon Budget of the
       United States: Carbon Pools and Flux Under Alternative Policy Options.  Report No. EPA/600/3-
       93/093. Washington, DC:  U.S. Environmental Protection Agency, May.

U.S. Department of Agriculture, Forest Service.  1990. An Analysis of the Timber Situation in the United
       States: 1989-2040, A Technical Document Supporting  the 1989 USDA Forest Service  RPA
       Assessment.  Haynes, R.W., Coordinator. General Technical Report No. RM-199. Fort Collins,
       CO: Rocky Mountain Forest and Range Experiment Station, December.

U.S. Department of Energy, Office of Policy, Planning, and Program Evaluation. 1994. The Climate
       Change Action Plan: Technical Supplement. Report No. DOE/PO-0011. Washington,  DC,
       March.

U.S. Department of Energy. 1991. National Energy Strategy, Technical Annex 2, Integrated Analysis
       Supporting the National Energy Strategy: Methodology, Assumptions, and Results. Report No.
       DOE/S-0086P. Washington, DC.

U.S. Department of State. 1994. Climate Action Report: Submission of the United States of America
       Under the United Nations Framework Convention on Climate Change. Washington, DC: U.S.
       Government Printing Office.
                                             63

-------
U.S. Environmental Protection Agency, Office of Policy, Planning, and Evaluation. 1994. Inventory of
       U.S. Greenhouse Gas Emissions and Sinks: 1990-1993. Report No. EPA 230-R-94-014.
       Washington, DC, September.

The White House. 1993. The Climate Change Action Plan. Washington, DC: U.S.  Government Printing
       Office, October.

Winnett, S., R.W. Haynes, and W.G. Hohenstein. 1993. "Economic Impacts of Individual Climate
       Change Mitigation Options in the U.S. Forest Sector." Climate Research 3:121-128.
                                             64

-------