1710
The impacts of changes in federal timber harvest
on forest carbon sequestration in western Oregon
Eun Ho Im, Darius M. Adams, and Gregory S. Latta
Abstract: This study examines the potential impacts of changes in federal timber harvest, acting through regional log markets, on the sequestration of carbon in forests and forest products in western Oregon. We construct a dynamic model of
the region’s log markets in which market prices, log consumption at mills, and timber harvests and timber inventories on
private, federal, and state forests are endogenous. Absent any policies regulating forest carbon sequestration, simulations
show that regional carbon flux in forests and forest products would gradually decline as federal harvest rises from recent
historical levels. If regional forest carbon flux were constrained to meet some minimum target, however, projections indicate that there would be opportunities for substituting carbon sequestration between federal and nonfederal lands through
coordination of harvests across ownerships. We find that relatively small reductions in average private harvest could offset
substantial losses of carbon flux on federal timberlands caused by increased federal harvest. One mechanism for achieving
the changes needed in private harvest to meet a regional carbon flux target would be a carbon tax/subsidy program or a
carbon offset market. For example, if federal owners offered timber for sale equal to the maximum sustainable level under
the Northwest Forest Plan, our analysis indicates that a carbon price of roughly $US 19 per tonne of carbon would be sufficient to induce private owners to undertake the harvest and management modifications necessary to maintain regional
forest carbon flux at its level in the early 2000s.
Résumé : Cette étude s’intéresse à l’impact que des changements dans la récolte de bois sur les terres fédérales, dont la
répercussion se fait sentir sur les marchés régionaux de grumes, peuvent avoir sur la séquestration du carbone dans les forêts et les produits forestiers dans l’ouest de l’Oregon. Nous avons construit un modèle dynamique des marchés de grumes
de la région dans lequel les prix du marché, la consommation de grumes dans les usines ainsi que les volumes de bois et
les inventaires de bois dans les forêts privées, les forêts fédérales et les forêts d’État sont endogènes. En l’absence de politiques régissant la séquestration du carbone forestier, les simulations montrent que le flux régional de carbone dans les forêts et les produits forestiers diminuerait graduellement à mesure que la récolte sur les terres fédérales augmente par
rapport au niveau des dernières années. Cependant, si l’on exigeait que le flux régional de carbone forestier atteigne un niveau quelconque minimum, les prévisions indiquent que la séquestration du carbone pourrait faire l’objet de substitutions
entre les terres fédérales et non fédérales en coordonnant la récolte parmi les propriétaires. Nous constatons que des réductions relativement faibles de la récolte provenant du secteur privé pourraient compenser les pertes substantielles de flux de
carbone dues à l’augmentation de la récolte sur les terres fédérales. Un programme de taxe ou subvention carbone ou un
marché d’échange du carbone serait un mécanisme qui permettrait de réaliser les changements nécessaires dans la récolte
du secteur privé pour atteindre un niveau souhaité de flux régional de carbone. Par exemple, si les propriétaires fédéraux
mettaient en vente un volume de bois égal au rendement soutenu maximum selon le « Northwest Forest Plan », notre analyse indique qu’un prix du carbone d’environ 19 $US la tonne serait suffisant pour inciter les propriétaires privés à entreprendre les modifications en termes d’aménagement et de récolte nécessaires pour maintenir le flux régional de carbone à
son niveau du début des années 2000.
[Traduit par la Rédaction]
Introduction
The Northwest Forest Plan (NWFP) (also called the
‘‘President’s Plan’’) was established in the mid-1990s at the
request of President Clinton to meet the need for late-successional forest habitat for certain endangered species while
providing some level of federal timber harvest to support
Pacific Northwest processing industries. In western Oregon,
the NWFP authorized annual harvests from federal lands of
approximately 600 million board feet (US Department of
Agriculture, Forest Service, and US Department of the Interior, Bureau of Land Management 1994a). This timber harvest level has never been implemented, however, due to
changes in agency policies and court challenges of federal
timber sales offerings by environmental and other groups.
Some industry and local government groups continue to
Received 19 March 2010. Accepted 12 May 2010. Published on the NRC Research Press Web site at cjfr.nrc.ca on 13 August 2010.
E.H. Im. International Cooperation Division, Korean Forest Service, Daejon, South Korea.
D.M. Adams1 and G.S. Latta. Department of Forest Engineering, Resources and Management, Oregon State University, Corvallis, OR
97331, USA.
1Corresponding
author (e-mail: [email protected]).
Can. J. For. Res. 40: 1710–1723 (2010)
doi:10.1139/X10-110
Published by NRC Research Press
Im et al.
urge federal agencies to raise cuts to NWFP levels, citing
growth in regional wood processing needs, employment benefits, and (most recently) as part of efforts to reduce fire
hazard.2 At the same time, the concerns raised about the impacts of federal harvest in the early 1990s, loss of wildlife
habitat, impaired fisheries production, and deterioration of
amenity values, remain important today, and continued harvest restrictions enjoy wide support.
As the United States moves toward a national policy of
controlling greenhouse gas emissions, a further concern
could potentially be added to this list: that increased harvest
would reduce rates of carbon sequestration in public forests,
particularly if cutting involves older stands with large carbon stores in both standing trees and downed woody debris
(Harmon et al. 1990; Schulze et al. 2000). Public forest
lands in the West are important carbon sinks with significant
rates of net carbon uptake. For the combined region of western Oregon and western Washington, Smith and Heath
(2004) estimated that net nonsoil carbon flux for all public
timberlands was roughly 200 million metric tons (tonnes) of
carbon (MMTC) between 1953 and 2000, nearly 10% of the
estimated net flux for all public timberlands in the coterminous United States over this period.
Projections of future carbon flux on forest lands in the
United States suggest that a substantial amount of additional
carbon could be sequestered on private and public ownerships (Smith and Heath 2004; US Environmental Protection
Agency 2005; Depro et al. 2008). Under a ‘‘current policies’’ scenario, the US Environmental Protection Agency
(2005) projected that private forest lands could be a net carbon sink absorbing about 66 MMTC/year by 2050, while
Depro et al. (2008) projected average annual carbon stock
change on US public lands at about 54 MMTC over this
same period. The work of Smith and Heath (2004) and Depro et al. (2008) also suggests that these public lands estimates are highly sensitive to assumptions about future
timber harvest and silvicultural practices.
This study examines the potential changes in regional carbon flux resulting from shifts in federal timber harvest in
western Oregon and identifies options for joint public–private strategies to control regional flux. To explore these issues, we consider two broad questions:
(1) How would changes in timber harvest on federal timberlands alter regional forest carbon flux under current policies? We develop a set of market scenarios in which timber
harvest from federal forests rises by stages from zero to the
highest volume that can be sustained within the bounds of
current federal forest land-use allocations and harvest guidelines. In these scenarios, the regional carbon flux impacts reflect only the interaction of ownerships in the timber market.
(2) Could timber harvests from federal and nonfederal
timberlands be coordinated in ways that reduce the carbon
flux impacts of changes in federal cut or even increase overall regional carbon flux? In these scenarios, we estimate the
1711
market clearing timber supplies from nonfederal and federal
lands subject to a set of alternative lower bounds on regional
forest carbon flux: a set of regional ‘‘carbon flux targets’’.3
To simulate these cases, we construct a dynamic model of
western Oregon log markets in which market prices and volumes together with harvest, growth, and inventory on both
private and public forests are explicitly modeled. In our
analysis, federal timber harvest varies with market demand
up to a maximum cut limit (the ‘‘volume offered for sale’’)
determined by agency policies and independent of market
behavior. Subject to this harvest bound, the selection of
stands to be harvested and modes of forest management
over time can be varied, as limited by the NWFP, so as to
alter the impacts on carbon flux. Thus, silvicultural regimes,
timber harvest timing, and growth are endogenous and potentially sensitive to alternative policy targets.
Past studies of public–private harvest interaction and associated carbon impacts have either (i) merged public and private timber inventories and timber harvest decisions as if
they shared common management, an approach that, in our
view, is too much at variance with current and past policy
structures governing federal lands management (e.g., Sedjo
and Lyon 1990; Sohngen and Mendelsohn 1998), or (ii)
looked only at the private response to a limited number of
fixed management and harvest options for public lands,
with no ability to consider management changes on public
lands in reaction to private shifts (Murray et al. 2004;
Adams and Latta 2005). The present study departs from
past work in two ways. First, the model developed here allows a systematic exploration of a wide range of alternative
federal harvest volume and forest management options to
determine the carbon flux implications on federal lands and
the simultaneous harvest and carbon flux responses on private lands. Second, given this capability, we can consider
how harvest and management choices on both ownerships
might vary in pursuit of a joint regional carbon flux objective.
Methods
Projection of future timber market activity and carbon sequestration involves five basic components: (i) initial inventory data, (ii) assumptions about the potential range of future
silvicultural regimes that might be applied to public and private lands, (iii) projections of future timber and carbon
yields under the several management regimes, (iv) assumptions about changes in timberland areas and land-use restrictions on private and public timberlands, and (v) a market
model describing current and future regional timber market
conditions. In this process, the market model projects future
timber harvest and prices based on initial inventory and
other assumptions, selects the management regimes to employ that are consistent with the objective, and updates
growing stock inventory and carbon storage over time.
2 See,
for example, the testimony of Tom Nelson on behalf of the American Forest and Paper Association before the Agriculture, Nutrition,
and Forestry Committee of the United State Senate on 26 June 2003 and the testimony of the American Forest and Paper Association
submitted to the House Appropriations Subcommittee on Interior, Environment, and Related Agencies on 16 March 2006 (www.afandpa.
org/) and the National Forest Counties and Schools Stabilization Act submitted by Douglas County in western Oregon (proposed safety net
solution is available from www.co.douglas.or.us/NFCS%20Stabilization%20Act%201-15-07.pdf).
3 Adams et al. (1999) employed a flux target approach in an analysis of national level carbon sequestration options on US private lands.
Published by NRC Research Press
REGEN + grow only
THIN HI + grow only
Matrix, AMA, RR
Matrix, AMA, RR
Matrix, AMA, RR
Matrix, AMA
THIN LO
THIN ME
THIN HI
Regeneration harvest
(REGEN)
LSR, AMA
Land-use allocation
LSR
Note: mbf, thousand board feet: a board foot is a measure of sawtimber volume (approximately 6.74 m3; Spelter 2004); 1 acre = approximately 0.405 ha; ‘‘volume percent’’ is the fraction of total volume
in stems at least 30 in. (76.2 cm) diameter at breast height (DBH).
additional practices after
additional practices after
additional practices after
THIN ME + grow only
Any period (no
thinning)
Any period (no
thinning)
Any period (no
thinning)
Any period (no
regeneration)
THIN LO + grow only
CT + grow only
Management intensity classes
Class
Grow only
Management action
Commercial thinning (CT)
Criteria
If >9 mbf/acre and volume percent (DBH ‡ 30
in.) <40%, remove 40% thinning across diameter classes (>7 in. and <36 in.)
If >9 mbf/acre and volume percent (DBH ‡ 30
inches) <60%, remove 40% thinning across
diameter classes (>7 in. and <36 in.)
If >9 mbf/acre, remove 20% thinning across
diameter classes (>7 in.)
If >10 mbf/acre, remove 40% thinning across
diameter classes (>7 in.)
If >10 mbf/acre, remove 60% thinning across
diameter classes (>7 in.)
If >12 mbf/acre and stand age ‡100 (or volume
percent (DBH‡30 in.) ‡60%), remove all trees
except 16 live trees/acre
Definitions of management practices based on NWFP Standards and Guidelines
Table 1. Management practices and management intensity classes used for public stands in the model.
Management intensity classes (MICs)
Regimes of silvicultural practices are applied to private,
state, and federal timberlands over the projection period, in
some cases more than one regime for stands harvested multiple times during the projection. For private lands, seven
MICs were defined for existing stands that are part of the
original inventory at the start of the projection. Stands established during the course of the projection could employ either natural or artificial regeneration and one of four
regimes of subsequent treatments (a total of eight options).
Details of the MICs for private timberlands are given in
Adams et al. (2002). We followed the same approach to the
MICs for state forests to be generally consistent with management guidance for these lands. Three partial cutting regimes were allowed in existing stands corresponding to
light, moderate, and heavy thinning, and regeneration harvest (clearcutting) can be applied to softwood stands.
For federal timberlands, the choice of management regime
is limited because all federal timberlands are allocated to
management areas (each with different management guidelines) by the NWFP. Categories include Congressionally Reserved area (CR) (e.g., wilderness areas), Late-Successional
Reserves (LSR), Adaptive Management Areas (AMA), Managed Late-Successional Areas (MLSA), Administratively
Withdrawn Areas (AW), Riparian Reserves (RR), and Matrix (the remaining area lying between the preceding allocations). Allowable management practices based on our
interpretation of the management standards and guidelines
(US Department of Agriculture, Forest Service, and US Department of the Interior, Bureau of Land Management
1994b) in each area are summarized in Table 1. The NWFP
guidelines provide general management directions, including
prohibited actions and spatial considerations of silvicultural
practices based on the desired future conditions of forests in
each area. They do not suggest specific silvicultural actions
or regimes. We have chosen relatively simple regimes, consistent with the guidelines that are suitable for our model
specification. There are two commercial thinning MICs for
softwood stands in LSR-related categories, with a different
regime where LSR and AMA overlap. Three partial cutting
regimes are defined for Matrix, AMA, and RR categories.
Timing applied
Initial period
Inventory
Timber inventory data for private and public forest lands
in western Oregon derive from the US Forest Service Forest
Inventory and Analysis (FIA) periodic forest survey (Azuma
et al. 2002). Because the inventories (collected between
1994 and 1998) were 9–12 years old, we updated the inventory data to a common 2005 starting point using a timber
harvest scheduling model. Harvest was constrained to match
actual historical cut at the county level and the areas clearcut and partial cut were constrained to match actual patterns
by year and owner at the half-state (western Oregon) level
based on data from the Oregon Department of Forestry.4
Federal timber harvest was assumed to occur in the areas
designated as ‘‘Matrix’’ and ‘‘Adaptive Management Areas’’
in the NWFP (see discussion of land classes in the following
section). Tree lists from the original plots were updated using a version of the ORGANON model for private and state
forest lands and the Forest Vegetation Simulator (FVS) for
federal lands (Hann et al. 1997; Dixon 2003).
additional practices after
Can. J. For. Res. Vol. 40, 2010
Any period (no additional practices after
thinning)
1712
Published by NRC Research Press
Im et al.
Regeneration harvest (clearcutting) is allowed only in softwood stands allocated to the Matrix and AMA designations.
Yield projection
Timber yields for private and state timberlands were generated using methods similar to those in Adams et al.
(2002). The individual tree simulation model ORGANON
(Hann et al. 1997) was used to model stand development
over time for each MIC in each stand. Tree lists from the
FIA data were input to ORGANON for existing stand types.
Regenerated stand tree lists were developed using the young
stand simulator SYSTUM1 (Ritchie 1993) for a range of site
classes for each ecological region and for Douglas-fir (Pseudotsuga menziesii (Mirb.) Franco) stands and other conifer
stands. Subsequent yields for all stands were developed
with ORGANON.
Inventory updates and yield estimates for each MIC on
federal forests were derived from the Forest Service FVS
stand projection system (Dixon 2003). Ingrowth for thinnings was assumed to occur naturally, with density that varied according to the number of trees removed by harvesting.
To mimic development of newly planted stands, the densities of regeneration after harvest were derived from averages for young stands from the FIA database by forest type
and ecoregion.
Carbon estimation
In our analysis, carbon storage in the forest ecosystem comprises five pools: live and dead trees, understory, forest floor,
woody debris, and soil. Carbon in standing trees was estimated
using tree lists projected by ORGANON and FVS and tree biomass equations. Stem biomass calculated from volume equations in the FIA was converted to stem carbon. Carbon in bark,
live and dead branches, and foliage was obtained from biomass equations multiplied by biomass–carbon conversion factors (Gholz et al. 1979; Ter-Mikaelian and Korzukhin 1997).5
Because most species do not have a root biomass equation, except for Douglas-fir, the amount of root carbon was approximated by the amount of aboveground carbon multiplied by
estimated ratios of above- to belowground biomass. Future biomass in snags is based on tree mortality, as predicted by the
individual tree simulation models, which is gradually oxidized
at a constant rate over time (Turner et al. 1995; Chen et al.
2001). Understory carbon by forest type was estimated as a
constant percentage of live tree carbon as suggested by the
US Environmental Protection Agency (2003) for the Pacific
Northwest region. Forest floor carbon is assumed to be 10%
of the carbon pools in the foliage and dead branches of the
live trees. Half of above- and belowground woody residues is
assumed to remain after logging and site preparation and is
gradually oxidized using constant decay rates (Edmonds
1987; Chen et al. 2001). Carbon in current snags and woody
debris is calculated from the FIA data. Estimated carbon in
mineral soil is based on Birdsey’s (1992) estimates and is, as
in most studies, assumed to be constant over time.
1713
Following Row and Phelps (1996) and Skog and Nicholson (2000), we estimated carbon in forest products by accounting for stocks and flows of carbon from forests to
wood and paper products in use, to dumps or landfills, and
to burning and emissions from decay. Carbon in western Oregon log market production and consumption was projected,
tracking each period’s wood harvest, import, and export
through to its final disposition. The timber allocated to sawmills, pulp mills, plywood plants, and other manufacturers is
divided into primary product and utilized/unutilized mill residue based on the Forest Service’s Timber Assessment projections (Haynes et al. 2007) and a statewide census of
Oregon’s forest products industry (Brandt et al. 2006). Carbon in solid wood and pulp products is transferred to several
end-use categories and the future carbon that remains sequestered in these products after disposal in dumps and
landfills is calculated. The current carbon stock in forest
products in use and landfills was estimated using the past
50 years of harvest data for western Oregon.
Land area changes and land-use allocation
In western Oregon, private owner groups have realized
small gains in timberland area since the late 1970s, for industry owners through transfers from other private owners,
and for other private owners through afforestation of land in
nonforest uses. State and federal timberland areas have been
relatively stable for several decades. In the absence of any
significant historical area trends, we assumed constant timberland area by owner during the projection period in this study.
Federal and state timberlands are divided into several
land-use categories according to existing forest plans. For
each federal stand (inventory plot), proportions of land-use
allocations designated by the NWFP were identified using
land-use allocation and Key Watersheds maps provided by
the Regional Ecosystem Office. Timberland classified as
CR and AW and inside Key Watersheds was excluded from
harvest. As a result, the total timberland area in this study
that may receive some form of harvest is 4.124 million acres
(54% of total US Forest Service and Bureau of Land Management timberlands). Of these areas, 45% (1.849 million
acres) is designated in the Matrix and AMA categories,
which may receive regeneration harvest and (or) thinning regimes. Riparian Reserves comprise 0.788 million acres and
are eligible for three types of partial cutting. About 1.487
million acres in LSR and LSR plus AMA lands could be
treated with some form of thinning to develop multistoried
stand structure. State-owned timberlands are also divided
into several use categories: reserved forests, multiresource
forest, and active forest with associated restrictions on harvest and management.
Market model
Future levels of timber harvest and prices are determined
by interactions among producers and consumers in the regional timber market. In this study, we construct an inter-
4 Oregon
annual timber harvest reports from 1986 to 2005 are available from the Oregon Department of Forestry website: egov.oregon.gov/
ODF/STATE_FORESTS/FRP/annual_reports.shtml.
5 We estimate carbon in trees from biomass equations, since individual tree information is available from ORGANON and FVS yield projections. In addition, it allows avoiding potential bias in using constant conversion of merchantable volume into total tree biomass
(Johnson and Sharpe, 1983).
Published by NRC Research Press
Regional carbon flux target is obtained from the AFC scenario.
Note: Carbon flux is the discounted sum of carbon flux over the projection periods (see eq. A8 in Appendix A).
a
Alternative regional carbon flux (ARCF): federal cut upper bound set (range: NFC to MSFC), private cut varies
with market and carbon constraint
Scenario acronym
Definition (minimum regional carbon flux)
a
Regional carbon flux same as baseline federal cut (BFC) case
BRCF
RCF@BFCka
Regional carbon flux same as k times the baseline federal cut (BFCk) cases
RCF@MSFCa
Regional carbon flux same as maximum sustainable federal cut (MSFC) case
BRCF+k%
k% additional regional carbon flux relative to the carbon flux in the baseline
federal cut (BRCF) case
Alternative federal cut (AFC): federal cut upper bound varies
by scenario, private cut varies with market, no constraints on
regional carbon flux
Scenario acronym
Definition (maximum federal cut level)
NFC
No federal cut
BFC
Recent average (2000–2004)
BFCk
k times the baseline (BFC)
MSFC
Maximum sustainable level under NWFP
temporal optimization model of the log market in western
Oregon similar to that described by Adams and Latta
(2007). Demand is derived from lumber and plywood production and log exports, all of which are sensitive to the delivered price of logs. The supply of logs from private forests
is based on land owners’ decisions about harvest timing and
silvicultural practices to optimize the value of their timber
investments given stand growth, interest rates, management
costs, and price expectations. Oregon law directs that state
lands are to be managed so as to maximize timber values
consistent with social and environmental values. Thus, state
owners are assumed to maximize net timber revenues subject to constraints on achieving future forest conditions described in the state forest plans. On federal lands, the
NWFP is intended to provide timber production at predictable levels while protecting and managing the long-term
health of forests, wildlife, and waterways. But while the
NWFP provides land-use allocations and management
guidelines, it is silent about the selection of specific areas
for harvest treatment within zones where harvesting may occur, and there is little objective information on how land
managers actually make these selections. To operationalize
this process, we assume that federal agencies choose areas
to harvest so as to minimize timber management costs (regeneration and subsequent activities) subject to the limitations of the NWFP (Table 1) and policy-based constraints
on the maximum level of harvest. Since timber management
costs are related to site conditions rather than current stand
conditions, this approach does not imply any simple standbased harvest ordering rule such as ‘‘oldest first’’.
A simplified mathematical outline of the regional log
market model is show in Appendix A. Market equilibria for
all periods in the projection are found where the present
value of consumers’ willingness to pay less the costs of timber management, log transport, and capacity maintenance
and expansion by regional log processors (eq. A1) is maximized (Samuelson 1952). Log demand (the price-dependent
function Pt(Qt, Kt) in eq. A1) shifts depending on product
prices, technology, nonwood costs, and capacity. The shortterm elasticities of demand for logs in the lumber and plywood sectors in western Oregon (at sample period means)
were estimated as –0.47 and –0.72, respectively (see further
discussion in Appendix A). Over time, the effective log demand elasticities will be larger, however, as lumber and plywood processing capacities (the measure of capital stock)
adjust. Processing capacity decisions (DK) are endogenous,
depending in part on equipment costs, depreciation, and interest rate. Capacity also forms a bound on regional log consumption (eqs. A1, A5, and A6). Changes in capacity shift
the log demand equations (intercept shift only since the demand relations are linear), altering the time path of demand
adjustment to any change in the market.
Regional log supply is derived from timber harvest from
private, state, and federal lands and net log imports from
other regions and must be at least as large as log demand
(eqs. A3 and A4). Inventories are characterized in terms of
areas in existing and regenerated stands, E and R, respectively, which are the areas of particular stands managed
under a specific harvest pattern and the management regimes applied to both (I). All existing stands must be allocated to some management regime (clearcut, partial cut, or
Can. J. For. Res. Vol. 40, 2010
Table 2. Definitions of the policy scenarios simulated in this study.
1714
Published by NRC Research Press
Im et al.
1715
Table 3. Comparison of estimates of nonsoil carbon stock density and forest carbon flux related to study area.
Carbon flux (gm–2year–1)
Carbon density (tonnes/ha)
Study
Heath et al. 2003a
Smith and Heath 2004b
Birdsey and Lewis 2003a
Law et al. 2004b,c
This studyb
Region; owners
Washington, Oregon,
California; all
Washington, Oregon; US
Forest Service
Washington, Oregon; OPUB
Oregon; all
Western Oregon; all
Western Oregon; all
Western Oregon; US Forest
Service
Western Oregon; OPUB
Forest
products
Live biomass
102.8
Dead biomass
45.6
180.8–188.8
70.0–72.3
133.0
82.3
164.1–196.8
150.3
205.3
53.0
42.4
38.4–49.5
35.7
44.9
31.5
19.7
161.7
34.0
22.5
Forest land
Forest
products
189.–226
232
21
26.4
.
a
Estimates for 1997.
Estimates for 2002; OPUB represents other federal lands and state represents nonfederal lands that are owned by state and local governments and other
nonfederal public owners.
c
Range of values for western Oregon ecoregions.
b
reserve). Newly created stands (by clearcutting) can occupy
no more area than was clearcut harvested over the projection
and may be either harvested again during the projection or
reserved from harvest (eq. A2). Stands managed under a partial cutting regime must be so designated at the start of the
projection and cannot be shifted to a clearcut regime during
the projection.
Federal cut is constrained by an ‘‘allowable cut’’ that
varies with harvest policies (eq. A7) used in the alternative
federal harvest scenarios associated with question 1 posed
in the Introduction. Regional forest carbon flux depends on
harvest and growth and may be constrained to meet some
minimum flux target (eq. A8) used in the carbon target scenarios related to question 2 in the Introduction.
Additional assumptions
Assumptions about future prices of processed products
and of labor and ‘‘other’’ variable inputs were derived from
projections in the 2005 Resource Planning Act timber assessment update (Haynes et al. 2007). Existing management
plans for all public forest lands require that they provide
sustainable and predictable production of forest products in
the future. This requirement is mimicked by imposing nondeclining flow restrictions on timber harvests and limitations
on period-to-period fluctuations of areas clearcut and
thinned. A multispecies habitat conservation plan has been
established by the State of Oregon to ensure that state forest
land management complies with the federal Endangered
Species Act (16 U.S.C 1531 et seq.). We approximate this
management objective by requiring that 50% of state forest
land be composed of complex stand structures (layered and
older forest structures) by 2035 rising to 60% by 2065.
Stand structure classes are defined by stand attributes as
suggested by Latta and Montgomery (2004).
Analysis of the impacts of changing federal harvest and
regional carbon flux
The impacts of changing federal timber harvest on forest
carbon flux and log markets in western Oregon are exam-
ined using sets of simulation scenarios. Table 2 presents definitions of the policy scenarios associated with the two
questions posed in the Introduction.
(1) Alternative federal cut (AFC): What are the potential
regional carbon flux changes resulting from various levels
of timber harvest on federal lands under current policies?
The concern here is to identify the trade-off between federal
timber harvest and regional carbon flux when no efforts are
made to modify the carbon flux impacts of harvest changes.
Future market behavior and carbon flux are projected in a
series of scenarios with rising maximum levels for federal
harvest. In the simplified model structure used in Appendix
A, this would correspond to rising levels of HFMAX in eq. A7
starting from zero (NFC), to the current or ‘‘baseline’’ federal harvest (BFC) computed as the average of recent historical levels, up to the highest harvest volume (MSFC) that
can be sustained within the limits of the NWFP. For each
timber harvest bound (first column in Table 2), HFMAX, the
model will project an associated total regional forest carbon
flux, TC
MIN, via eq. A8. We expect that federal forest carbon
flux will fall as federal cut rises, but the overall impact on
total regional flux given possible substitution of federal for
private/state cut in the log market is unclear a priori.
(2) Alternative regional carbon flux (ARCF): Could timber harvests from federal, private, and state lands be coordinated in ways that reduce the regional carbon flux impacts
of changes in federal cut or even increase overall regional
carbon flux? The flux levels found in the AFC simulations
above were used to define lower bounds for the ARCF scenarios. For instance, in Table 2, ‘‘BRCF’’ represents the regional carbon flux associated with the current or ‘‘baseline’’
cut level (BFC), while ‘‘RCF@BFCk’’ is the regional carbon flux when maximum federal allowable cut is k times the
baseline level (BFCk). The market model was simulated
using these fluxes as regional carbon flux minimums or targets (TC
MIN in eq. A8). In these scenarios, the total regional
carbon flux is constrained to be at least TC
MIN but the levels
and mix of federal, private, and state harvest are free to
vary. For a given set of scenarios, maximum federal cut is
Published by NRC Research Press
1716
Can. J. For. Res. Vol. 40, 2010
Table 4. Average annual harvest and annual carbon sequestration in forests and forest products by owner under the
alternative federal harvest (AFC) and regional carbon flux (ARCF) scenarios in western Oregon, 2005–2065 (see
Table 2 for definition of acronyms).
Average annual harvest (1000 m3)
Annual carbon sequestration (1000 tonnes)
Policy scenario
Private
Federal
State
Total
Private
AFC
NFC
BFC
BFC2
MSFCa
21499
21518
21492
21543
0
943
1885
6217
2913
2912
2912
2912
24411
25373
26290
30671
ARCF targets with maximum federal cut set at MSFC
RCF@MSFC
21543
6217
2912
30671
RCF@BFC2
21463
6139
2923
30525
BRCF
21430
6137
2924
30490
BRCF+10%
21013
6167
2920
30101
BRCF+20%
20668
6158
2902
29728
BRCF+80%
16578
5839
2861
25278
Federal
State
Totalb
2460
2463
2460
2470
3889
3718
3516
2378
2140
2139
2139
2139
12381
12213
12008
10879
2470
3357
3546
4725
5901
12959
2378
2525
2540
2564
2582
2754
2139
2249
2251
2284
2336
2449
10879
12024
12230
13465
14711
22054
a
Maximum sustainable federal cut (MSFC) is about 6.6 times the current or baseline federal cut (BFC).
Total includes annualized carbon sequestration, 3892 thousand tonnes, on federally owned reserved areas.
b
fixed at some level but actual cut can fall below the maximum as needed to most efficiently meet the carbon flux target. In effect, the several ownerships are ‘‘jointly’’ managed
so as to meet the total regional carbon flux target with minimum market impact. Maximum federal cut can range from
zero (NFC) to MSFC. For each upper bound on federal cut,
there is a set of ARCF or ‘‘flux target’’ scenarios corresponding to the flux levels found for the AFC cases (see Table 2) with flux levels below that found in the base case
(BRCF). There is a further set in which we expand the flux
by fixed percentages above the baseline flux (BRCF+k%).
For any given upper bound on federal cut, regional harvest is expected to fall as the flux target rises, but given
(downward) flexibility in federal cut and the significant differences between private, state, and federal inventories, the
resulting changes in the mix of cut and carbon sequestration
by ownership are difficult to forecast.
Results and discussion
An initial concern is how well our forest carbon estimates
for the current period compare with results from other studies of the Pacific Northwest region and Oregon. Table 3
presents our estimates of current nonsoil carbon densities
and carbon flux for forest lands and forest products together
with results from previous studies. Heath et al. (2003),
Smith and Heath (2004), and Birdsey and Lewis (2003)
used carbon accounting approaches that derive carbon mass
though adjustments of total merchantable inventory volumes. Their estimates of carbon density for live biomass
are generally lower than ours because their study regions include large areas of land in eastern Oregon, eastern Washington, and California with lower site quality and lower
timber stocking than in western Oregon. Law et al. (2004),
in contrast, used remote sensing methods and eddy flux
data from a network of sampling points to produce a carbon
budget for the forested region of western Oregon alone. Our
estimate of average carbon density for all owners is slightly
below the ecoregion range reported by Law et al. (2004),
while our average carbon flux is slightly higher than the
upper limit of their flux range.
Baseline projection
The baseline (scenario BFC) represents a case where future timber harvest from federal timberlands can be no
larger than the average between 2000 and 2004, and there
are no changes in regulatory policies for federal and nonfederal lands. Baseline simulation results showed that the total
estimated nonsoil carbon stock of western Oregon forests
and forest products would gradually rise to 1977 MMTC by
2065, a 45% increment. At the owner level, nonsoil carbon
stocks by 2065 would increase by 21% on private lands,
55% on federal lands, and 82% on state lands.
The projected baseline net carbon increase on federal
lands amounts to some 442 MMTC by 2065, representing
an average annual rate of 7.4 MMTC during the projection
period. This is about 15% of the annual carbon flux on all
US federal lands projected by Depro et al. (2008) for a similar time period. Smith and Heath (2004) estimated that the
public forests of the Pacific Northwest Westside (western
Oregon and Washington) could accumulate an additional
400 MMTC by 2040. Our estimate of carbon increase on
federal lands in western Oregon alone is about 330 million
tonnes by 2040. The lower increment in Smith and Heath
(2004) may be due, in part, to their use of a somewhat
higher federal harvest rate. Within the federal land aggregate, our study projects that the carbon stock in dead biomass will double by 2065 because of high mortality. In
contrast, carbon stored in forest products transferred from
federal forests will drop to roughly half of the current level.
Changing federal cut with and without regional carbon
flux constraints
The AFC scenarios portray market behavior under a series
of fixed maximum federal harvest levels beginning at zero
and rising to the maximum sustained yield level. Simulation
Published by NRC Research Press
Im et al.
1717
Fig. 1. Timber harvest and carbon sequestration trade-offs under the alternative federal cut (AFC) and regional carbon flux (ARCF) scenarios for different levels of maximum federal cut in western Oregon, 2005–2065. See Table 2 for simulation acronyms and definitions.
results for average annual regional harvest and forest carbon
sequestration are presented in the upper portion of Table 4.
Significantly, our model projects that average private and
state harvests would change little in the face of expanded
federal cut. As federal harvest rises in the scenarios, the
model projects that regional processing capacity will also increase, damping price effects and limiting public–private
harvest substitution. As expected, average regional carbon
sequestration steadily declines as harvest rises. Carbon sequestration on federal lands, however, does not appear to be
highly sensitive to changes in annual federal harvest. The
6.6-fold increase in federal maximum cut from the base
(current) level to the maximum sustainable level (MSFC)
leads to a 36% reduction in federal sequestration. Regional
cut rises by 21%, while regional sequestration falls by 11%.
The ARCF scenarios simulate a range of gradually rising regional carbon sequestration targets, or minimum sequestration
rates, allowing harvests on all ownerships to vary as dictated
by their specific carbon sequestration characteristics and market demands. The results shown in the lower part of Table 4
are for the case where maximum federal cut is set at MSFC.
Similar tabulations could be developed for each lower level of
maximum cut down to zero. Unlike the AFC cases, sequestration in these simulations is fairly sensitive to harvest changes
because most of the adjustment is taken in private harvest. For
example, a 4% reduction in private cut (from flux at the maximum federal sustained yield level to flux 20% above baseline
rates) leads to a 39% increase in private sequestration. Over
this same range, regional cut falls by 3%, while regional sequestration rises by 35%. In these scenarios, federal harvest is
relatively stable, remaining near the maximum sustained yield
level until the largest sequestration target (Table 4).
Carbon–harvest substitution
The potential for harvest and carbon sequestration tradeoffs based on the AFC and ARCF scenarios is illustrated in
Fig. 1. The relatively flat AFC curve reflects the limited response of sequestration to changing cut in these cases (as
noted above), while the ARCF curves (whatever the maximum federal cut) are more sensitive to cut reductions. The
bases for these differential responses lie in the differences
between inventories on federal and private lands. The age
class distribution of private lands is heavily concentrated in
younger stands with low carbon stocks but high rates of carbon sequestration. The harvestable (unreserved) portions of
federal lands are older with slower volume growth and carbon flux rates. Carbon densities (tonnes per hectare) are also
higher than on private lands but not as high as densities on
the reserved (nonharvestable) federal areas. In the AFC scenarios, 70% of federal harvests come from stands 100 years
and older and 90% from stands with carbon flux rates of
1%/year or less. As maximum federal cut rises from zero in
these scenarios, the loss in carbon flux is limited. This result is
obtained despite the explicit inclusion of losses due to carbon
losses in pools of large and small woody debris on the forest
floor in our estimates of carbon flux in existing forest stands.
In the carbon target (ARFC) scenarios, imposing higher
regional sequestration minimums forces owners to lengthen
rotation age. On private lands, this has relatively large impacts on net sequestration, given the high carbon growth
rates of the younger stands that dominate the age structure.
Changes in private harvest concentration by age and carbon
flux rate can be seen in Figs. 2 and 3. Rotations lengthen
and cut shifts into stands with lower carbon flux rates. On
federal lands, NWFP guidelines allow few opportunities to
Published by NRC Research Press
1718
Can. J. For. Res. Vol. 40, 2010
Fig. 2. Percentage of average annual private harvest by stand age class under the baseline and alternative regional carbon flux (ARCF)
scenarios in western Oregon, 2005–2065.
alter the concentration of harvest by stand carbon flux rate,
so few adjustments are made in federal cut in the ARCF
runs (until the highest target rates).
The relation of AFC and ARCF curves in Fig. 1 suggests
that, ignoring all other resource values on federal lands, the
region could realize the current or higher levels of forest
carbon sequestration and higher harvest levels. This would
entail expanding maximum federal harvest and reducing private cut. If federal harvest were allowed to expand as far as
maximum sustained yield, regional harvest and flux could
shift from the baseline point (BFC) to some point on the
ARCF curve in Fig. 1. The welfare effects of such a shift
on market participants would be varied, as illustrated in Table 5 by changes from BFC to points in the ARCF range.
Simply maintaining the current flux level (BRCF) while revising harvest across owners would raise total regional surplus by 23%, with nearly equal gains in surplus by
consumers (mills) and federal timber sellers (sales receipts)
but modest losses by private and other timber suppliers.
If expansion of federal harvest beyond current levels is
not an option, regional flux could still be increased by
lengthening private rotations and reducing private and total
regional cut. This option is illustrated in Fig. 1 by the longdashed line labeled ARCF (maximum federal cut equal to
current cut). In fact, a whole sheaf of trade-off curves exists
similar to the cases with maximum and current federal cut.
Each would start at a point along the AFC line, with a federal harvest upper bound somewhere between zero (NFC)
and maximum sustained yield, and slope upward and to the
left as cut is reduced (examples indicated by the shortdashed lines in Fig. 1).
Operational issues
Private cut
If it were possible to impose a regional carbon target for a
given maximum federal cut level, obtaining the optimal mix
of lands to harvest on federal ownerships might be possible
by administrative directive. But how would compliance be
obtained on private lands? One possibility is the development of a market for carbon offsets: a carbon price. Under
the ARCF scenarios, the shadow price of the carbon sequestration constraint (eq. A8) represents the marginal cost of
changing the regional carbon flux target by one unit (Hoen
and Solberg 1994). It is the marginal welfare cost to market
participants (consumers and producers) for sequestering an
additional unit of carbon in forests and forest products in
western Oregon under the regional market system. It is also
the price of carbon that would induce private forest owners
to manage their lands to optimally meet the carbon target.
The solid line in Fig. 4 represents the marginal costs of carbon sequestration in western Oregon forests under the ARCF
scenarios with maximum federal cut at MSFC. Carbon flux
increments of up to 6 MMTC/year could be obtained with a
carbon price lower than $US 50/tonne C. In Fig. 1, these
would be flux increments moving along the solid ARCF curve
beginning at the MSFC point. In this case, raising regional
flux to the base level (BRCF) would require an increase of
about 1.4 MMTC/year. From the data of Fig. 4, this could be
achieved with a carbon price of about $US 18.70/tonne C.
The dashed line in Fig. 4 illustrates the marginal costs of
flux increments in the case where maximum federal cut is
set at current levels. In the context of Fig. 1, these would
Published by NRC Research Press
Im et al.
1719
Fig. 3. Percentage of average annual private harvest by carbon growth rate class under the baseline and alternative regional carbon flux
(ARCF) scenarios in western Oregon, 2005–2065.
Fig. 4. Shadow prices (marginal market welfare costs) of incremental carbon sequestration in forests and forest products under the regional
carbon flux (ARCF) scenarios in western Oregon, 2005–2065.
be increases beginning at the point BFC and moving upward
to the left on the dashed ARCF curve. Costs here are
slightly higher for increments beyond 1 MMTC/year than in
the previous case.
Leakage
This study provides only a partial equilibrium analysis of
western Oregon timber markets. Changes in forest policy af-
fecting western Oregon timber supply may cause changes in
carbon flux in other regions as a result of interactions in
product markets, a process termed ‘‘leakage’’ (Aukland et
al. 2003; Murray et al. 2004; Sohngen and Brown 2004).
For the combinations of public and private cut that lead to
both more harvest and higher carbon flux (the area above
and to the right of the construction lines in Fig. 1), leakage
would entail a reduction in harvest and expansion of net carPublished by NRC Research Press
1720
Can. J. For. Res. Vol. 40, 2010
Table 5. Market surplus in log markets by owner under the alternative federal cut
(AFC) and regional carbon flux (ARCF) scenarios (with maximum federal cut at
MSFC) in western Oregon, 2005–2065 (see Table 2 for definition of acronyms).
Market surplus (2002 $US million)
Producer surplus
Policy scenario
Consumer surplus
Federal
Othersa
Total surplus
AFC
- NFC
- BFC
- BFC2
- MSFC
1009
1072
1142
1468
0
68
135
403
1815
1804
1793
1744
2824
2944
3070
3615
ARCF with maximum federal cut at MSFC
- RCF@MSFC
1468
403
- RCF@BFC2
1468
401
- BRCF
1468
401
- BRCF+10%
1466
402
- BRCF+20%
1430
402
- BRCF+80%
995
400
1744
1734
1731
1703
1672
1314
3615
3603
3599
3571
3504
2708
a
‘‘Others’’ includes private and state forest owners and log importers.
bon flux in other regions through substitution in product
markets. The net effect for, say, North America would likely
be an increase in forest sector carbon flux. Increased product
output from western Oregon would be offset in part by reduced product output and harvest in other regions. This
would change the mix of carbon increments stored as products and in forests. But since the carbon flux in western Oregon would be at least as large as the base, while cut in
other regions would be no higher than the base (due to substitution), net North American flux should rise.
For federal harvest limits that yield regional harvests below the base (as in the dashed ARCF curve in Fig. 1 with
maximum federal cut equal to current cut), reductions in
western Oregon harvest would stimulate some increase in
harvests in other regions, again through substitution in product markets.6 The net effect on forest carbon flux in the
broader market region is unclear in these cases without resorting to a market model of the larger region.
Summary and conclusions
This study employed a model of the log market and forest
carbon stocks in western Oregon to examine the impacts of
alternative levels of federal timber harvest on regional forest
carbon sequestration. We also analyzed potential trade-offs
between public and private harvest when forest owners, including federal agencies, made joint efforts to achieve alternative levels of regional carbon flux. In a series of
simulations with changes only in federal cut, results show
that regional carbon flux from forests and forest products
would gradually decline as federal harvest rises. Higher federal cut leads to lower federal carbon flux. Substitution of
federal timber for private and state harvests was found to be
limited, even with federal cut at its maximum sustainable
level, resulting in modest changes in carbon flux on nonfed-
eral lands. As federal cut rises, processing capacity expands
to absorb the additional harvest volume and price impacts
are modest. This reflects both the model’s 5-year time step
(which allows significant capital stock changes) and its assumption of optimal (present net worth maximizing) capital
stock adjustment. Producer and consumer surplus impacts
(relative to the current policy base case) are consistent with
these market changes: log processors realize an increment in
surplus through expanded log consumption and a slight drop
in log price, federal agencies receive higher timber sales
revenue, and private and state forest land owners are persistent, although modest, losers. Rising federal timber supply
leads to reduction in regional carbon sequestration in a
nearly linear fashion as simulated harvest increases from
zero to the maximum sustainable level.
In contrast, the alternative regional carbon flux scenarios
show that, with coordination of harvests between federal
and nonfederal lands, higher regional carbon flux could be
maintained even as regional harvest increases. Carbon flux
on private lands is highly sensitive to harvest change, while
on federal and state lands, it is relatively insensitive. The
trade-off curve between timber harvest and regional carbon
flux was found to be concave to the origin (Fig. 1); the carbon flux/harvest trade-off rises as harvest expands. These
trade-offs are brought about by substituting timber harvest
and carbon sequestration between the federal and private
owner groups. Reductions in carbon flux on federal lands
can be offset by modifications in the form and extent of cut
on private lands. This entails a cut reduction by lengthening
rotation ages and shifting the concentration of harvest within
an ownership toward stands with lower carbon densities and
rates of carbon growth.
Implementing a harvest coordination program as envisioned in the carbon target scenarios may entail significant
6 This
case, the effects of movement along the ARCF curve with current cut as the federal maximum, is the traditional case and is the
subject of the study by Im et al. (2007).
Published by NRC Research Press
Im et al.
issues of management and control. If federal harvest were
flexible, our results suggest that similar targets could also
be achieved through a carbon tax/subsidy system or a carbon
offsets market. The shadow price of the carbon target constraint for the case of a baseline flux target (BRCF), for example, indicates that a carbon price of just $US 18.7/tonne
C ($US 5.1/tonne CO2) would be sufficient to induce owners to modify harvests and management activities.7
In considering these results, one must recognize that we
have ignored all other resource values save those for carbon
uptake and timber production and our accounting of welfare
changes derives exclusively from the regional log market.
We have assumed that western Oregon federal harvest
would be allowed to rise above current levels to reach more
efficient harvest – C flux combinations. The options considered are hypothetical and intended to demonstrate potential
trade-offs. Although there are current proposals to raise federal cut from several groups,8 it is not our intent to advocate
for any particular viewpoint in this debate. Assessment of
the impacts of carbon flux policies on other resources, impacts that were central in the development of the NWFP
more than 15 years ago, should be a key task of future research.
Acknowledgements
This research was supported by the College of Forestry,
Oregon State University, and the Pacific Northwest Research Station, USDA, Forest Service, Portland, Oregon.
Conclusions and opinions are strictly those of the authors.
References
Adams, D.M., and Latta, G.S. 2005. Costs and regional impacts of
restoration thinning programs on the national forests in eastern
Oregon. Can. J. For. Res. 35(6): 1319–1330. doi:10.1139/x05065.
Adams, D.M., and Latta, G.S. 2007. Timber trends on private lands
in western Oregon and Washington: a new look. West. J. Appl.
For. 22: 8–14.
Adams, D.M., Alig, R.J., McCarl, B.A., Callaway, J.M., and
Winnett, S.M. 1999. Minimum cost strategies for sequestering
carbon in forests. Land Econ. 75(3): 360–374. doi:10.2307/
3147183.
Adams, D.M., Schillinger, R.R., Latta, G.S., and Van Nalts, A.K.
2002. Timber harvest projections for private land in western
Oregon. Research Contribution 37. Forest Research Laboratory,
Oregon State University, Corvallis, Ore. fcg.cof.orst.edu/rc/rc37.
pdf.
Aukland, L., Costa, P.M., and Brown, S. 2003. A conceptual framework and its application for addressing leakage: the case of
avoided deforestation. Clim. Policy, 3: 123–136.
Azuma, D., Bednar, L., Hiserote, B., and Veneklase, C. 2002. Timber resource statistics for western Oregon, 1997. U.S. For. Serv.
Resour. Bull. PNW-RB-237.
Birdsey, R.A. 1992. Carbon storage and accumulation in United
States forest ecosystems. U.S. For. Serv. Gen. Tech. Rep. WO-59.
Birdsey, R.A., and Lewis, G.M. 2003. Carbon in U.S. forests and
wood products, 1987–1997: state-by-state estimates. U.S. For.
Serv. Gen. Tech. Rep. NE-310.
1721
Brandt, J.P., Morgan, T.A., Dillon, T., Lattman, G.J., Keegan, C.E.,
and Azuma, D.L. 2006. Oregon’s forest products industry and
timber harvest, 2003. USDA Forest Service, Pacific Northwest
Research Station, Portland, Ore.
Chen, H., Harmon, M.E., and Griffiths, R.P. 2001. Decomposition
and nitrogen release from decomposing woody roots in coniferous forests of the Pacific Northwest: a chronosequence approach. Can. J. For. Res. 31(2): 246–260. doi:10.1139/cjfr-31-2246.
Depro, B., Murray, B., Alig, R., and Shanks, A. 2008. Public
land, timber harvest and climate mitigation: quantifying carbon sequestration potential of U.S. public timberlands. For.
Ecol. Manag. 255(3–4): 1122–1134. doi:10.1016/j.foreco.
2007.10.036.
Dixon, G.E. 2003. Essential FVS: a user’s guide to the Forest Vegetation Simulator. Internal Report. U.S. Department of Agriculture, Forest Service, Forest Management Service Center, Fort
Collins, Co.
Edmonds, R.L. 1987. Decomposition rates and nutrient dynamics in
small-diameter woody litter in four forest ecosystems in Washington, U.S.A. Can. J. For. Res. 17(6): 499–509. doi:10.1139/
x87-084.
Gholz, H.L., Grier, C.C., Campbell, A.G., and Brown, A.T. 1979.
Equations for estimating biomass and leaf area of plants in the
Pacific Northwest. Res. Pap. 41. Forest Research Laboratory,
Oregon State University, Corvallis, Ore.
Hann, D.W., Hester, A.S., and Olsen, C.L. 1997. ORGANON
user’s manual. Edition 6.0. Department of Forest Resources,
Oregon State University, Corvallis, Ore.
Harmon, M.E., Ferrell, W.K., and Franklin, J.F. 1990. Effects on
carbon storage of conversion of old-growth forests to young forests. Science, 247(4943): 699–702. doi:10.1126/science.247.
4943.699. PMID:17771887.
Haynes, R.W., Adams, D.M., Alig, R.J., Ince, P.J., John, R., and
Zhou, X. 2007. The 2005 RPA timber assessment update. U.S.
For. Serv. Gen. Tech. Rep. PNW-GTR-699.
Heath, L.S., Smith, J.E., and Birdsey, R.A. 2003. Carbon trends in
U.S. forestlands: a context for the role of soils in forest carbon
sequestration. In The potential of U.S. forest soils to sequester
carbon and mitigate the greenhouse effect. Edited by J.M. Kimble, L.S. Heath, R.A. Birdsey, and R. Lal. CRC Press, Boca Raton, Fla. pp. 35–45.
Hoen, H.F., and Solberg, B. 1994. Potential and economic efficiency of carbon sequestration in forest biomass through silvicultural management. For. Sci. 40(3): 429–451.
Im, E., Adams, D.M., and Latta, G.S. 2007. Potential impacts of
carbon taxes on carbon flux in western Oregon private forests.
For. Policy Econ. 9: 1006–1017.
Johnson, W.C., and Sharpe, D.M. 1983. The ratio of total to merchantable forest biomass and its application to the global carbon
budget. Can. J. For. Res. 13(3): 372–383. doi:10.1139/x83-056.
Latta, G., and Montgomery, C.A. 2004. Minimizing the cost of
stand level management for older forest structure in western
Oregon. West. J. Appl. For. 19(4): 221–231.
Law, B.E., Turner, D., Campbell, J., Sun, O.J., Van Tuyl, S., Ritts,
W.D., and Cohen, W.B. 2004. Disturbance and climate effects
on carbon stocks and flux across western Oregon, USA. Glob.
Change Biol. 10(9): 1429–1444. doi:10.1111/j.1365-2486.2004.
00822.x.
Murray, B.C., McCarl, B.A., and Lee, H.C. 2004. Estimating leak-
71
tonne of carbon = 3.67 tonnes of CO2.
for example, the Association of O&C Counties proposal at www.co.douglas.or.us/NFCS%20Stabilization%20Act%201-15-07.pdf. A
similar option was proposed by the Forest Service Employees for Environmental Ethics (www.fseee.org).
8 See,
Published by NRC Research Press
1722
age from forest carbon sequestration programs. Land Econ.
80(1): 109–124. doi:10.2307/3147147.
Ritchie, M.W. 1993. User’s guide for SYSTUM-1: a simulator
of growth trends in young stands under management in California and Oregon. U.S. For. Serv. Gen. Tech. Rep. PSWGTR-147.
Row, C., and Phelps, R.B. 1996. Wood carbon flows and storage
after timber harvest. In Forests and global change. Vol. 2. Forest
management opportunities for mitigating carbon emissions. Edited by R.N. Sampson and D. Hair. American Forests, Washington, D.C. pp. 27–58.
Samuelson, P.A. 1952. Spatial price equilibrium and linear programming. Am. Econ. Rev. 42: 283–303.
Schulze, D.E., Wirth, C., and Heimann, M. 2000. Managing forests
after Kyoto. Science, 289(5487): 2058–2059. doi:10.1126/
science.289.5487.2058. PMID:11032555.
Sedjo, R.A., and Lyon, K.S. 1990. The long-term adequacy of
world timber supply. Resources for the Future, 1616 P Street
NW, Washington, D.C.
Skog, K.E., and Nicholson, G.A. 2000. Carbon sequestration in
wood and paper products. In The impact of climate change on
America’s forests: a technical document supporting the 2000
USDA Forest Service RPA assessment. Edited by L.A. Joyce
and R. Birdsey. USDA Forest Service, Rocky Mountain Research Station, Fort Collins, Co. pp. 79–88.
Smith, J.E., and Heath, L.S. 2004. Carbon stocks and projections
on public forestlands in the United States, 1952–2040. Environ.
Manag. 33(4): 433–442. doi:10.1007/s00267-003-9101-x. PMID:
15453397.
Sohngen, B., and Brown, S. 2004. Measuring leakage from carbon
projects in open economies: a stop timber harvesting project in
Bolivia as a case study. Can. J. For. Res. 34(4): 829–839.
doi:10.1139/x03-249.
Sohngen, B., and Mendelsohn, R. 1998. Valuing the impact of
large-scale ecological change in a market: the effect of climate
change on U.S. timber. Am. Econ. Rev. 88(4): 686–710.
Spelter, H. 2004. Converting among log scaling methods: Scriber, International, and Doyle versus cubic. J. For. 102(4):
33–39.
Ter-Mikaelian, M.T., and Korzukhin, M.D. 1997. Biomass equations for sixty-five North American tree species. For. Ecol.
Manag. 97(1): 1–24. doi:10.1016/S0378-1127(97)00019-4.
Turner, D.P., Koerper, G.J., Harmon, M.E., and Lee, J.J. 1995. A
carbon budget for forests of the conterminous United States.
Ecol. Appl. 5(2): 421–436. doi:10.2307/1942033.
US Department of Agriculture, Forest Service, and US Department
of the Interior, Bureau of Land Management. 1994a. Final supplemental environmental impact statement on management of
habitat for late-successional and old-growth forest related species within the range of the northern spotted owl. Vol. 1. Portland, Ore. [Irregular pagination]
US Department of Agriculture, Forest Service, and US Department
of the Interior, Bureau of Land Management. 1994b. Record of
decision for amendments to Forest Service and Bureau of Land
Management planning documents within the range of the northern spotted owl. [Place of publication unknown; 74 p . plus attachment A: standards and guidelines]
U.S. Environmental Protection Agency. 2003. Inventory of U.S.
greenhouse gas emissions and sinks, 1990–2001. U.S. Environmental Protection Agency, Washington, D.C.
U.S. Environmental Protection Agency. 2005. Greenhouse gas mitigation potential in U.S. forestry and agriculture. U.S. Environmental Protection Agency, Washington, D.C.
Can. J. For. Res. Vol. 40, 2010
Appendix A. Simplified representation of
market model structure
To reduce notational complexity, references to product
type (lumber, plywood) and the details of log import and export have been suppressed. Maximize
½A1
T 1
X
ð1 þ rÞt
R
Pt ðQt ; Kt ÞdQt km Kt
t¼0
R
ku DKt Mt ðSt Þ þ Nt ðZt ÞdZt R
þ r1 ð1 þ rÞT PT ðQT ; KT ÞdQT
R
km KT MT ðST Þ þ NT ðZT ÞdZT subject to
X
½A2
ait ðEti ; Rit ; Iti Þ Ai for i ¼ P; F
t
½A3
St ¼ HtP ðEtP ; RPt ; ItP Þ þ HtF ðEtF ; RFt ; ItF ; polÞ þ Zt for 8 t
½A4
Qt St for 8 t
½A5
Kt ð1 dÞKt1 DKt 0 for 8 t
½A6
lKt Qt for 8 t
½A7
F
for 8 t
HtF ðEtF ; RFt ; ItF ; polÞ HMAX
½A8
T1
X
ð1 þ rÞt ½DCtP ðEtP ; RPt ; ItP Þ þ DCtF ðEtF ; RFt ; ItF Þ
t¼0
C
þ FWðSt Þ TMIN
where t and T are time period and the length of the projection period, respectively, r is the discount rate, EPt , RPt , EFt ,
and RFt are the areas of existing and regenerated stands in
private/state (P) and federal (F) ownerships at period t, IPt
and IFt represent the arrays of management regimes (or management intensities) available on private/state and federal
lands at period t, aPt (EPt , RPt , IPt ) and aFt (EFt , RFt , IFt ) are land
area accounting equations that control the relationship between areas of existing (E) and regenerated (R) stands and
their allocation to management regimes (I) at period t and
insure that no more than the available land base (A) is used
for private/state and federal ownerships, respectively, AP and
AF are the timberland areas of private/state and federal ownerships, Qt is log consumption in milling processes at period
t, Kt and DKt are capital stock (represented as milling capacity) and capital stock change at period t, Pt(Qt, Kt) is the inverse derived demand curve for logs by processing
industries at period t, QT is average log consumption in
milling processes at the postprojection period, d is the capital depreciation rate, l is the maximum capital stock utilization rate, km and ku are the unit costs of maintaining capital
stock and purchasing new capital, respectively, St is log supply from private/state and federal forests and net import at
period t, Mt(St) is the total cost of timber supply St at period
Published by NRC Research Press
Im et al.
t including costs for timber harvesting, transportation, and
management on private/state and federal lands, HPt () and
HFt () give the volumes of timber harvested on private/state
and federal lands as functions of land area in existing and
regenerated stands and area allocations to the various management regimes at period t, ‘‘pol’’ represents the effects of
federal timber harvest policies (as in the NWFP standards
and guidelines) in specifying the timber that can be harvested and how it can be cut, Nt() is the inverse net log
trade equation (export log demand less import log supply),
Zt is net log trade from/to other regions (offshore and other
US regions) at period t, HFMAXis the upper bound on federal
timber harvest, only employed in the alternative federal harvest (AFC) scenario, DCPt and DCFt are sets of functions that
compute the net carbon flux in forestlands and forest products on private/state and federal forest lands at period t,
FW(St) is carbon in fuelwood used in milling processes at
period t counted as a fossil fuel offset, and TC
MIN is an optional lower bound on regional forest carbon sequestration,
only employed in the alternative regional carbon flux
(ARCF) scenario.
The objective function involves the customary maximization of market surplus, in this case the present value of producers’ plus consumers’ surpluses net of capacity expansion,
harvesting, transport, and other costs. Terminal conditions
involve the valuation of ending inventory (at the end of the
projection) assuming a perpetual harvest stream based on the
final inventory volume (assuming full regulation and using
von Mantel’s formula). Price varies according to the final
period’s demand relation. We also attempt to avoid the influence of ending conditions on our results by developing
long projections but employing only the earliest portion in
the policy analysis.
Log demand equations for the lumber and plywood sectors were derived from profit function analyses using annual
data on outputs, factor consumption, and prices over the period 1970–1998 (details of an earlier application are described in Adams et al. 2002). We employed a normalized,
restricted quadratic functional form with global curvature
properties (convexity) imposed. Input categories included
logs, labor, capital, and ‘‘all other’’; technology was represented by a time trend. Capital stock was treated as quasifixed and measured by estimated maximum physical production capacity. Parameter estimates were obtained by iterative
3SLS, treating log price as endogenous.
1723
Log imports to western Oregon (from western Washington and other locations) were assumed to have an import
supply elasticity of 1. Export log demands (offshore) were
assumed to have a fixed elasticity of –0.4. The net demand
relation and the net export flow are represented as the inverse function Nt and the variable Zt, respectively, in eq. A1.
Endogenous variables are EP, RP, EF, RF, and DK. After
linearization of P(Q, K), the problem can be solved as a linear program for a given sequence of DK, yielding values for
E and R and a new series of DK. This new series is used in
a second solution and so on in an iterative fashion until the
DK stabilizes within some tolerance. This Gauss–Seidel approach eliminates the need to solve the overall system eqs.
A1–A8 as a nonlinear programming problem (see Montgomery et al. 2006). Equations A7 and A8 are only used in certain scenarios: eq. A7 is used in alternative federal cut
(AFC) scenario and eq. A8 in the alternative regional carbon
flux (ARCF) scenario. Linear programming solutions were
obtained with the GAMS system employing the CPLEX
solver.
Projections are made for 100 years with a 5-year time
step. In this paper, however, we examine only the first
60 years of the projection as the relevant policy period and
to avoid undue influence of the terminal conditions. The
simulations reported here use a real discount rate of 6% for
all owners. Further details can be found in Schillinger et al.
(2003) and Im et al. (2007) who employed similar models,
including discussion of approaches used to recognize nontimber objectives for nonindustrial private forest owners.
References
Adams, D.M., Schillinger, R.R., Latta, G.S., and Van Nalts, A.K.
2002. Timber harvest projections for private land in western
Oregon. Research Contribution 37. Forest Research Laboratory,
Oregon State University, Corvallis, Ore. fcg.cof.orst.edu/rc/rc37.
pdf.
Im, E., Adams, D.M., and Latta, G.S. 2007. Potential impacts of
carbon taxes on carbon flux in western Oregon private forests.
For. Policy Econ. 9: 1006–1017.
Montgomery, C.A., Latta, G.S., and Adams, D.M. 2006. The cost
of achieving old-growth forest structure. Land Econ. 82: 240–
256.
Schillinger, R.R., Adams, D.M., Latta, G.S., and Van Nalts, A.K.
2003. An analysis of future private timber supply potential in
western Oregon. West. J. Appl. For. 18: 1–9.
Published by NRC Research Press