ArticlePDF Available

Effects of Management on Carbon Sequestration in Forest Biomass in Southeast Alaska

Authors:

Abstract and Figures

The Tongass National Forest (Tongass) is the largest national forest and largest area of old-growth forest in the United States. Spatial geographic information system data for the Tongass were combined with forest inventory data to estimate and map total carbon stock in the Tongass; the result was 2.8 ± 0.5 Pg C, or 8% of the total carbon in the forests of the conterminous USA and 0.25% of the carbon in global forest vegetation and soils. Cumulative net carbon loss from the Tongass due to management of the forest for the period 1900–95 was estimated at 6.4–17.2 Tg C. Using our spatially explicit data for carbon stock and net flux, we modeled the potential effect of five management regimes on future net carbon flux. Estimates of net carbon flux were sensitive to projections of the rate of carbon accumulation in second-growth forests and to the amount of carbon left in standing biomass after harvest. Projections of net carbon flux in the Tongass range from 0.33 Tg C annual sequestration to 2.3 Tg C annual emission for the period 1995–2095. For the period 1995–2195, net flux estimates range from 0.19 Tg C annual sequestration to 1.6 Tg C annual emission. If all timber harvesting in the Tongass were halted from 1995 to 2095, the economic value of the net carbon sequestered during the 100-year hiatus, assuming 20/Mg C, would be20/Mg C, would be 4 to 7 million/y (1995 US dollars). If a prohibition on logging were extended to 2195, the annual economic value of the carbon sequestered would be largely unaffected (7 million/y (1995 US dollars). If a prohibition on logging were extended to 2195, the annual economic value of the carbon sequestered would be largely unaffected (3 to $6 million/y). The potential annual economic value of carbon sequestration with management maximizing carbon storage in the Tongass is comparable to revenue from annual timber sales historically authorized for the forest.
Content may be subject to copyright.
Effects of Management on Carbon
Sequestration in Forest Biomass in
Southeast Alaska
Wayne W. Leighty,
1
*
Steven P. Hamburg,
1
and John Caouette
2
1
Center for Environmental Studies, Brown University, P.O. Box 1943, Providence, Rhode Island 02912, USA;
2
Regional Office,
Wildlife, Fisheries, Ecology and Watershed, US Forest Service, P.O. Box 21628, Juneau, Alaska 99802-1628, USA
ABSTRACT
The Tongass National Forest (Tongass) is the largest
national forest and largest area of old-growth forest
in the United States. Spatial geographic informa-
tion system data for the Tongass were combined
with forest inventory data to estimate and map
total carbon stock in the Tongass; the result was
2.8 ± 0.5 Pg C, or 8% of the total carbon in the
forests of the conterminous USA and 0.25% of the
carbon in global forest vegetation and soils.
Cumulative net carbon loss from the Tongass due
to management of the forest for the period 1900–95
was estimated at 6.4–17.2 Tg C. Using our spatially
explicit data for carbon stock and net flux, we
modeled the potential effect of five management
regimes on future net carbon flux. Estimates of net
carbon flux were sensitive to projections of the rate
of carbon accumulation in second-growth forests
and to the amount of carbon left in standing bio-
mass after harvest. Projections of net carbon flux in
the Tongass range from 0.33 Tg C annual seques-
tration to 2.3 Tg C annual emission for the period
1995–2095. For the period 1995–2195, net flux
estimates range from 0.19 Tg C annual sequestra-
tion to 1.6 Tg C annual emission. If all timber
harvesting in the Tongass were halted from 1995 to
2095, the economic value of the net carbon
sequestered during the 100-year hiatus, assuming
$20/Mg C, would be $4 to $7 million/y (1995 US
dollars). If a prohibition on logging were extended
to 2195, the annual economic value of the carbon
sequestered would be largely unaffected ($3 to
$6 million/y). The potential annual economic value
of carbon sequestration with management maxi-
mizing carbon storage in the Tongass is comparable
to revenue from annual timber sales historically
authorized for the forest.
Key words: carbon sequestration; geographic
information system; climate change; forest
management; Alaska.
I
NTRODUCTION
Concern over rising levels of atmospheric carbon
dioxide, a primary greenhouse gas (GHG), has gi-
ven impetus to the construction of global carbon
budgets. Forest carbon dynamics are a key com-
ponent of these budgets. Although the Kyoto
Protocol of the UN Framework Convention on
Climate Change provides for a potentially active
and regulated market in Certified Emission
Reduction credits (CERs) for some types of forest
management, implementing such a program has
been controversial, and as of 2006 the United
States has not ratified the Kyoto Protocol. Quan-
tifying sources and sinks of carbon and the fluxes
resulting from forest management is essential for
the accurate estimation of national emissions and
transparent functioning of a CER market that
could help a country meet GHG emission reduction
targets.
Received 8 March 2005; accepted 14 December 2005; published online
17 November 2006.
*Corresponding author; e-mail: WayneLeighty@alumni.brown.edu
Current address: P.O. Box 20993, Juneau, Alaska 99802, USA
Ecosystems (2006) 9: 1051–1065
DOI: 10.1007/s10021-005-0028-3
1051
Terrestrial vegetation and soil represent impor-
tant sources and sinks of atmospheric carbon
(Watson and others 2000), with land-use change
accounting for 24% of net annual anthropogenic
emission of GHGs to the atmosphere (Prentice and
others 2001). Managing these terrestrial carbon
stocks to mitigate future climate change requires
information on global and national carbon budgets.
Specifically, the management of public lands rep-
resents a policy challenge, because there is often a
mandate to consider multiple uses, including car-
bon storage or reduced emissions. Consequently,
estimating the potential economic value of the
carbon held in these lands, and the impacts of
management on carbon stocks, may become an
important part of managing public lands.
It is likely that CERs would be allocated based on
the change in total carbon stock caused by a shift in
forest management. Consequently, quantifying net
carbon flux under varied management regimes and
establishing a ‘‘business as usual’’ baseline are key
to planning for future uses of public lands.
We chose to study the carbon implications of
forest management of the Tongass National Forest
(Tongass) in southeast Alaska for several reasons.
First, the Tongass is the largest national forest in
the United States, and it is part of the largest intact
old-growth temperate rainforest in the world
(USDA Forest Service 2005). Second, few esti-
mates of terrestrial carbon pools include Alaska,
and we are aware of no estimates of net carbon
flux that include the Tongass (Turner and others
1995; R. A. Birdsey personal communication).
Based on studies of similar ecosystems in the US
Pacific Northwest, however, it is reasonable to
assume there is a large net carbon flux due to
harvesting in the Tongass (Harmon and others
1990; Smithwick and others 2002). Third, the
dearth of information about carbon flux in the
Tongass has prevented inclusion of the economic
value of carbon storage in the development of
forest management policies for the Tongass. Eco-
nomic value provides a common metric for com-
parison of the relative merits of carbon
management with other goods and services pro-
vided by the forest. Finally, knowledge about the
effects of management regimes on net carbon flux
in the Tongass will help define the relative
importance of the management of these federal
lands on GHG emissions in the United States.
Commercial timber harvesting began in the
Tongass in the early 20
th
century, and harvest
intensity increased in 1954 after the granting of
two 50-year timber contracts to large pulp mills
(Ketchikan Pulp Corporation and the Alaska Pulp
Corporation). In the 1990s, the timber volume
harvested from the Tongass declined as a result of
the closure of these two pulp mills. There was a net
loss to the Tongass timber program in 1998 of about
$29 million on $6.5 million in timber sales (USDA
Forest Service 2001).
The research reported in this study was designed
to assess Tongass carbon stocks in 1995, historic net
carbon flux from the Tongass, effects of future
management regimes on net carbon flux, and the
economic value of any net carbon sequestration
resulting from possible future management
regimes.
In this research, existing (1995) and historic
carbon stocks of the Tongass were estimated by
integrating geographic information system (GIS)
data with forest inventory data. Then this spatially
explicit model was used with accretion data from
permanent plots to examine the effects of five fu-
ture management regimes on net carbon flux for
the period 1995–2195.
M
ETHODS
The 70,000-km
2
Tongass National Forest lies
within the Pacific Northwest coastal temperate
rainforest biome, with average annual precipita-
tion of 150–560 cm, average winter temperatures
of )1to 10C, and average summer temperatures
of 10to 21C (Nowacki and others 2001). Gla-
ciers covered most of the region 14,000–20,000 y
bp and are now found in some valleys (Nowacki
and others 2001). Stretching 800 km along the
southeast coast of Alaska, the Tongass includes
22,000 islands with forest, muskeg, alpine mea-
dow, rock, fresh water, and ice (Nowacki and
others 2001; Everest and others 1997). Twenty
percent of the area of the Tongass is rock and ice,
12% is densely vegetated forestlands, 43% is
moderately vegetated forestlands, and 25% is
wetlands (USDA Forest Service 2000). The forest
composition of the Tongass in 1995, based on
species frequency in forest inventory data, was
43% Western hemlock (Tsuga heterophylla), 19%
Alaska yellow cedar (Chamaecyparis nootkatensis),
16% mountain hemlock (Tsuga mertensiana), 9%
Sitka spruce (Picea sitchensis), 7% western red ce-
dar (Thuja plicata), 5% lodgepole pine (Pinus con-
torta), and 1% other species (USDA Forest Service
1995b). In the 1970s, over 2000 km
2
(3%) of the
Tongass came under the control of Alaska Native
Corporations as a result of the Alaska Native
Claims Settlement Act. These lands were excluded
from this study because they lack comprehensive
forest inventory data.
1052 W. W. Leighty and others
Estimate of Existing Carbon Stocks
Calculation of Carbon Stocks at Sample Plots across the
Tongass. Carbon stocks were calculated for each of
the USDA Forest Service 1995 Forest Inventory
Assessment (FIA) Southeast Alaska Grid Inven-
tory’s 2000 systematic sample plots using data from
these plots (USDA Forest Service 1995b). Data on
live and dead vegetation (including diameter,
height, and species), downed woody debris, and
soil (including thickness of Oi, Oe, and Oa hori-
zons) were collected at each sampling plot (USDA
Forest Service 1995b) (see Appendix 1 at <http://
www.springerlink.com>).
We used these data to quantify the following
seven carbon pools for each FIA sampling plot: (a)
trees, (b) saplings/seedlings, (c) standing dead
wood, (d) coarse woody debris (CWD) (average
diameter more than 7.62 cm), (e) small woody
debris (SWD) (average diameter less than 7.62 cm
and large-end diameter more than 2.5 cm), (f)
understory vegetation, and (g) soil.
Allometric equations were used with tree diam-
eter and height data to estimate biomass (Mg/ha)
(see Appendix 2 at <http://www.springer-
link.com>). For species with more than one suit-
able allometric equation, biomass was estimated
using equations resulting in both the lowest and
highest biomass estimates (see Sensitivity Analy-
sis). To address the need to use most of the equa-
tions beyond the range of data from which they
were created, three-dimensional surface plots were
created to confirm consistent behavior of the
equations (for example, no inflection points) over
the range of diameter at breast height (dbh) and
heights to which they were applied. Additionally,
the total amount of carbon in trees larger than the
allometric equation bounds was estimated in our
sensitivity analysis. Root-to-shoot ratios for conif-
erous forests (with the exception of Pinus sylvestris,
a European species) range from 15% to 26%, so
belowground biomass was assumed to be 20% of
aboveground biomass (Santantonio 1977; Cairns
and others 1997; Hamburg and others 1997).
Additionally, belowground biomass was calculated
with the range 15%–26% of aboveground biomass
in our sensitivity analysis. Carbon was assumed to
account for 50% of tree biomass (Hamburg and
others 1997).
Standing dead biomass was calculated with the
same methods used for living trees, but with a decay
factor (0%–100% depending on the extent of decay
and component of the tree) (see Appendix 3 at
<http://www.springerlink.com>). Likewise, the
same allometric equations were used to calculate the
amount of carbon in seedlings and saplings (dbh less
than 2.5 cm and 2.5 to dbh 12.5 cm, respectively).
The amount of carbon in CWD was calculated
using FIA methods ((K. L. Waddel) public com-
munication 2001, An application of line intersect
sampling to estimate attributes of coarse woody
debris in resource inventories, USDA Forest Service
Pacific Northwest Research Station, Forest Sciences
Laboratory) (see Appendix 4 at <http://www.
springerlink.com>). The amount of carbon in SWD
was calculated with the methods described by
Brown (1974) (see Appendix 4 at <http://
www.springerlink.com>).
Understory biomass was calculated using the fo-
liar cover-to-biomass relationships developed in
Alaska by Yarie and Mead (1988). By aggregating
understory species described by Yarie and Mead
into the general taxonomic categories used in the
FIA, we calculated a species-weighted biomass
constant for each FIA category. Biomass in under-
story vegetation was then calculated by multiplying
these constants by foliar percent cover data from
the FIA horizontal/vertical (HV) subplot data. Bio-
mass estimates for each layer described in the FIA
HV data were summed to yield total understory
carbon stocks (Mg C/ha) for each FIA plot.
Soils data from the FIA Grid Inventory were
inadequate for accurately estimating soil carbon in
southeastern Alaska because only the top 50 cm
were sampled, but organic horizons alone are often
much deeper (Alexander and others 1989). Con-
sequently, total soil carbon in organic and mineral
horizons was calculated by applying the soil-cate-
gory carbon stocks developed for the Tongass by
Alexander and others (1989) to each of the more
than 800 soil management units (SMU) in the
Tongass (USDA Soil Conservation Service 1992a,
1992b; 1994; D. V. D’Amore personal communi-
cation 2001). Alexander and others used soil
samples and pedon descriptions to estimate aver-
age organic carbon stock (kg C/m
2
) for 10 gen-
eral soil categories in the Tongass (see Appendix
5 at <http://www.springerlink.com>). The SMU
scheme defines soil profiles, with the area of each
SMU mapped in polygons in a GIS database (GIS
polygon data define areas with defined attributes).
We began by grouping each SMU into the soil
categories described by Alexander and others
(1989). Then each SMU was assigned the carbon
stock given by Alexander and others for its associ-
ated category. When an SMU was intermediate to
two soil categories, it was assigned to the category
with a lower carbon stock to ensure a conservative
carbon estimate. Finally, total soil carbon in the
Tongass was calculated by multiplying the carbon
Carbon Sequestration in Southeast Alaska Forest 1053
stock assigned to each SMU by its total area. In the
10% of the Tongass where soil type has not been
mapped, mostly wilderness areas, soil carbon stock
was assumed to be the spatially weighted average
of all soil types. Total soil carbon in the Tongass was
also calculated from FIA soil pit data (see Sensitivity
Analysis).
Creation of Spatially Explicit Land-Cover Types and
Carbon Stock Estimates. Existing USDA Forest Ser-
vice GIS data (Figure 1) were combined using the
computer software ArcInfo 380 New York Street
Redlands, CA 92373-8100 (Environmental Systems
Research Institute; Workstation ArcInfo, copyright
1982–2002, ver. 8.0). A decision tree (Figure 1)
was applied to the resulting Complete Coverage for
the Tongass using SAS (SAS Institute Inc; 100 SAS
Campus Drive, Cary, NC 27513-2414 SAS System
for Windows, copyright 1989, 1996, release 6.12)
to define Total Carbon Polygon Types (TCPT) and
Aboveground Carbon Polygon Types (ACPT) based
on polygon attributes.
The decision tree uses available polygon attri-
butes to predict polygon types with varying
aboveground and belowground carbon stocks. For
example, an unharvested, productive spruce–
hemlock forest with high volume and size class
(ACPT 18) contains greater aboveground carbon
stocks than a harvested, productive forest with low
volume and size class (ACPT 23) (Figure 1).
Next, the polygons in the Complete Coverage
were aggregated based on their TCPT (370 polygon
types) and ACPT (40 polygon types) designation.
Figure 1. Decision tree delimiting polygon types with different carbon stocks. Ovals represent Aboveground Carbon
Polygon Types (ACPTs). These 40 polygon types exist for each of 11 soil-type categories, for a total of 370 Total Carbon
Polygon Types (TCPTs). Pattern-coded diamonds indicate data sources used in differentiating among polygons. Dotted lines
divide the figure into four general classes of ACPTs for ease of interpretation. MBF, million board feet (2360 m
3
); SMU, soil
associations and complexes; NFCON, nonforested conditions; FPROD, expected annual growth; VOLC, timber volume;
SSIZEC, dominant timber size; YR_CUT, year of timber harvest; FTYPE, general forest type; SLPCLS, slope gradient;
HYDRIC, hydric and nonhydric soil conditions; ASPECT-CODE, slope aspect.
1054 W. W. Leighty and others
Polygon slivers caused by the aggregation processes
in ArcInfo (defined as polygons with area less than
0.4 ha and perimeter/area ratio greater than 1) were
merged with their largest neighboring polygon.
Finally, the location of each FIA plot was asso-
ciated with a TCPT. The aboveground and below-
ground carbon stocks for each TCPT (Mg C/ha)
were then calculated by averaging the carbon
stocks for all FIA plots in the TCPT. The total carbon
stock in the Tongass was calculated by multiplying
the carbon stock for each TCPT by its area and
summing all TCPTs.
Projecting Net Carbon Flux
Equations were constructed to model carbon
accretion in aboveground biomass after harvesting
(Figure 2). Forest inventory data from 272 perma-
nent ‘‘growth and yield’’ plots from throughout the
Tongass were used to estimate biomass accumula-
tion over the first 100 years of regrowth (DeMars
2000). The area-weighted average aboveground
carbon stock of all old-growth commercial forest
ACPT was used to approximate the carbon stock of
forests more than 350 years old (assumed to be in
equilibrium) because prior research suggests it can
take 350 years for forests in southeast Alaska to
reach old-growth equilibrium (Janisch and Harmon
2002). We addressed the lack of data on biomass of
stands 100 to 350 years old by employing two car-
bon accretion models for 500 y of forest growth: a
polynomial (y¼91012 x53108x4þ4
105x30:0209x2þ4:6459x;R2¼0:8727) and
an asymptotic (y¼105x40:0027x3þ0:2078x2
1:0021x;R2¼0:9531). Comparison between
these two models enabled us to test the sensitivity of
flux estimates to the uncertain shape of this accu-
mulation curve.
Pools of CWD were assumed to increase after
harvest by 40% of the preharvest aboveground
standing biomass (estimated from FIA data) due to
stumps and slash left on site, and then decline with
decomposition (Sampson and Hair 1996). Carbon
stocks in the soil before and after harvest were as-
sumed to be unchanged due to lack of data
informing us otherwise.
Past net carbon flux, since 1900, was based on
historic harvest volumes. We split the harvest his-
tory in the Tongass into two time periods, 1900–54
and 1955–95, because the rate of timber harvest
increased dramatically in 1954 with the initiation
of two long-term timber contracts (USDA Forest
Service 1995a). Because nearly all timber harvest-
ing in the Tongass has involved clear-cutting, we
assumed that this harvest method would continue
in the future. Future net carbon flux was modeled
for the following five forest management regimes:
(a) no timber harvesting, regrowth of secondary
forest, and equilibrium in unharvested areas (a
lower bound for harvest intensity); (b) harvesting
of all forested lands on 100-year rotations (an up-
per bound for harvest intensity); (c) harvesting of
all forested lands on 200-year rotations (used to
examine the impact of harvest rotation period); (d)
harvesting of all lands currently available for har-
vest (exclusion of existing roadless areas) on 200-
year rotations (represents an approximation of
‘‘business as usual’’); and (e) harvesting of all lands
Figure 2. Carbon accretion
curves for aboveground live
biomass. Filled diamonds
represent data from permanent
plots; open diamonds are the
area-weighted average of old-
growth Aboveground Carbon
Polygon Types (ACPTs). The
solid line shows the best-fit
polynomial model of carbon
accretion; the dashed line is the
asymptotic accretion curve.
Variable site quality (site index)
causes divergence among
permanent plot data.
Carbon Sequestration in Southeast Alaska Forest 1055
currently available for harvest (exclusion of exist-
ing roadless areas) on 100-year rotations (used to
examine the impact of harvest rotation period).
Current land-use designations (USDA Forest Ser-
vice GIS coverage LUD99) were used to identify
areas available for harvest, and projected harvests
were spread evenly across available land.
Forest regrowth was assumed to follow the bio-
mass accretion models described above, with the
amount of carbon in a specific polygon dependent
on stand age and precut carbon stocks. For the
modeling of past net carbon flux, the total carbon
stock in 1900 was calculated by assuming that all
polygons were unharvested in 1900 and assigning
carbon stocks to harvested polygon types equal to
their unharvested equivalents (Figure 1). We allo-
cated the total net historic flux (difference between
carbon stock in 1900 and 1995) between the time
periods 1900–54 and 1955–95 in proportion to the
volume of timber cut in each period.
To estimate net carbon flux associated with
harvesting, we calculated the forest products
stream, the amount of carbon left on site as slash
and stumps, and the amount of carbon sequestered
annually in secondary growth at annual time steps.
Net annual carbon flux from the Tongass was cal-
culated as the total amount of carbon leaving the
forest less regrowth and does not include carbon
storage in forest products. Carbon storage in forest
products was included in estimates of net annual
carbon flux to the atmosphere, assuming that 60%
of the aboveground living biomass is merchantable
and the rest is left on site as slash and stumps
(Sampson and Hair 1996) (Figure 3). Historically,
roughly half of the merchantable volume entered
the sawtimber production process, whereas the
other half entered the pulpwood production
process (Warren 1999).
We assumed that 90% of the carbon in sawtim-
ber products was emitted to the atmosphere over
75 years (assuming an exponential release pattern),
and that the corresponding figures were 50 years
for pulpwood products, and 100 years for slash and
stumps left on site after harvesting (Skog and
Nicholson 1998). The CWD and SWD present prior
to harvesting was assumed to linearly lose half its
carbon in the 50 years after harvesting, accounting
for decreased deadwood formation in the early
stages of secondary growth. These carbon pools
were then increased to their preharvest stocks over
the next 200 years.
Aboveground carbon stocks after harvesting
were assumed to be equal to those in polygons
defined as forested, productive, low-volume, har-
vested areas with seedlings/saplings (ACPT 23) in
one set of model runs and to equal zero in another
(see Sensitivity Analysis).
Conversion of Net Carbon Flux to
Monetary Units
Current estimates of the economic value of carbon
in potential emissions trading markets vary widely,
from $5 to $125 Mg
)1
C (Weyant 2000); in this
analysis, we assumed a market value of $20 Mg
)1
C
for avoided emissions or sequestered carbon. We
did not apply a discount factor or temporal varia-
tion in this value, so all monetary values are in
1995 US dollars. Leakage, the possibility of offset-
ting increases in emissions associated with
increased harvest elsewhere, was not considered in
estimating the economic value of different man-
agement scenarios.
Sensitivity Analysis
To test the influence of assumptions required for
the analysis described above, we carried out sensi-
tivity analyses involving the following issues: the
selection of allometric equations, the use of allo-
metric equations for trees outside their specified
ranges, estimation of soil carbon, the shape of
biomass accretion curves, old-growth biomass of
cut-over lands, and postharvest carbon stocks.
Using the results of specific sensitivity analyses,
upper- and lower-bound estimates of net carbon
Figure 3. Product/waste flows for the southeast Alaska
timber industry. The timing of carbon flux to the atmo-
sphere varies among pathways. Percentages refer to the
proportion of the total carbon impacted by harvesting in
each product/waste. Figure modified from Sampson and
Hair (1996).
1056 W. W. Leighty and others
flux were calculated. These bounds indicate the
potential impact on our estimates of these key
sources of uncertainty, but they do not include all
possible sources of uncertainty. As such, the sen-
sitivity analysis cannot be considered an uncer-
tainty analysis capable of providing absolute
bounds on our estimates.
Tongass carbon pools were estimated using allo-
metric equations resulting in the lowest and high-
est biomass estimates for all species (Table 1).
Similarly, the importance of carbon in trees larger
than the size specified for the allometric models
employed was examined by calculating the total
amount of carbon in these trees.
Carbon in CWD at one FIA plot was an outlier
(more than twice the next nearest measurement);
therefore it was excluded from calculation of our
best estimates of carbon in CWD for this ACPT.
We included this high value in our calculations
during the sensitivity analyses to verify its relative
insignificance.
In addition to the calculation of soil carbon from
GIS SMU data described above, total soil carbon
was calculated from FIA soil pit data (thickness of
soil horizons) using carbon-density estimates
(Mg/m
3
) for each soil horizon (Alexander and
others 1989). For each FIA soil pit, horizon thick-
nesses were multiplied by their associated carbon
density estimate, as given by Alexander and others,
to estimate carbon stock. These carbon stock esti-
mates were used to estimate the carbon stock for
each SMU, which were then multiplied by the total
area of each SMU to calculate total soil carbon stock
in the Tongass. The total amount of soil carbon in
areas lacking soil GIS data was estimated, with both
methods, to gauge the size of this uncertain carbon
pool.
In calculating our upper- and lower-bound
carbon pool and net flux estimates, belowground
biomass was calculated using the upper (26%)
and lower (15%) bounds of applicable published
root-to-shoot ratios.
The time periods for 90% carbon emission from
the saw timber, pulp products, and slash pools were
both doubled and halved to gauge the influence of
these rates on the shape of projected net carbon
flux curves.
Net carbon fluxes were modeled using both
asymptotic and polynomial biomass accretion
curves (Figure 2). Net carbon flux was also calcu-
lated using mean and 95% confidence limits (CL)
for carbon stock estimates for each ACPT.
In the no-harvesting scenario, there was uncer-
tainty as to the long-term biomass accumulation on
cut-over lands. For example, will ACPT 7 eventu-
ally reach the carbon stock of ACPT 8 or 10 (Fig-
ure 1)? To test the sensitivity of net flux projections
to the assumed precut carbon stock, the model was
run assuming biomass accumulation to a carbon
stock of the most similar ACPT, as well as to the
carbon stock of a related ACPT with the highest
timber volume.
The carbon stock in aboveground standing bio-
mass of ACPT 23 was used as an estimate of the
amount of carbon present immediately after har-
vesting. However, this ACPT is defined as con-
Table 1. Carbon Pools in the Tongass National Forest in 1995
Model Runs
Carbon Pool (Pg) 1 2 3 4
Roots 0.12 0.12 0.12 0.04
Soil 1.86 1.86 1.86 1.86
Total aboveground 0.87 0.85 0.82 0.38
Trees 0.42 0.42 0.53 0.18
Seedlings/saplings 0.16 0.16 0.05 0.03
Dead Snags 0.09 0.09 0.10 0.04
CWD 0.18 0.16 0.12 0.12
SWD 0.00 0.00 0.00 0.00
Understory 0.02 0.02 0.02 0.02
Total (+ 95% CI) 2.85 (0.51) 2.83 (0.48) 2.80 (0.51) 2.28 (0.40)
CWD, coarse woody debris; SWD, small woody debris; CI, confidence interval.
Six model runs were made using the following combinations of allometric equations and assumptions to quantify the sensitivity of estimation to necessary assumptions: Run 1
used allometric equations predicting low carbon contents, did not include willow or birch, and included a CWD outlier. Run 2 used allometric equations predicting low carbon
contents, included willow and birch, and did not include a CWD outlier. Run 3 used allometric equations predicting high carbon contents, included willow and birch, and did
not include a CWD outlier. Run 4 used allometric equations predicting low carbon contents, included willow and birch, and did not include a CWD outlier or trees with dbh
greater than specified for each allometric equation.
Carbon Sequestration in Southeast Alaska Forest 1057
taining forest composed of seedlings and saplings
(86 Mg C/ha), which suggests that between 5 and
15 years have elapsed since harvesting in these
areas. Consequently, net flux projections were also
performed assuming zero carbon stocks in above-
ground standing biomass after harvesting, a clear
underestimate of aboveground living biomass on
recently clear-cut lands.
We did not explore the effects of varying our
assumptions about the forest products industry (for
example, the proportion of biomass used for mer-
chantable products or the ratio of sawtimber to
pulp production) in the calculation of upper and
lower bounds in our sensitivity analysis. Altering
these assumptions does influence the shape of our
projections of net carbon flux to the atmosphere
(Figure 4) but does not impact the magnitude and
was therefore not amenable to quantification in a
sensitivity analysis. Changing these assumptions
essentially hastens or delays carbon emission to the
atmosphere depending on whether more carbon is
entering product streams with longer or shorter life
spans. More detailed examination of this effect is
beyond the scope of this paper.
R
ESULTS
Evaluation of our spatially explicit carbon stock
estimates suggests that they are a realistic repre-
sentation of forest structure. Comparison of GIS
carbon stock coverages to aerial photographs
showed a correlation between observable transi-
Figure 4. Past and potential future aggregate net carbon flux between the Tongass and atmosphere (excluding soils). AD
Aggregate net carbon flux between the Tongass and the atmosphere with each management scenario, re-zeroed in 1995.
Asymptotic carbon accretion in secondary growth is assumed in A and B; polynomial carbon accretion in secondary
growth is assumed in C and D. Carbon stock in standing aboveground biomass after harvesting is assumed to be equal to
zero in B and D; carbon stock in standing aboveground biomass after harvesting is assumed to be equal to 86 Mg C/ha in A
and C. The total carbon stock in the Tongass was estimated to be 2.83 Pg in 1995. Negative aggregate net flux indicates
carbon emission from the Tongass; positive aggregate net flux indicates carbon accumulation in the Tongass.
1058 W. W. Leighty and others
tions in forest characteristics and mapped carbon
densities.
The creation of carbon stock polygons resulted in
a limited number of distinct and unique landscape
units. Twelve ACPTs account for over 90% of the
area of the Tongass, and 10 of them account for
86% of the total carbon (Figure 5). Polygon types
with few FIA sample plots have large uncertainty in
carbon stock estimates, but they represent small
land areas and contribute very little to the total
carbon stock. The 17 ACPT with less than five FIA
plots represent 2% of the area of the Tongass and
1% of the total carbon, whereas each of the
10 ACPTs that combine for 86% of the total carbon
stock of the Tongass have between 43 and 312 FIA
sample plots each.
The aboveground carbon stocks in each ACPT
correspond with qualitative descriptions of the
areas. Unharvested high volume old-growth forest
(ACPT 18), for example, has over five times the
aboveground carbon stock of a muskeg meadow
(ACPT 5) (Figures 1 and 5). The influence of soil
carbon, however, complicates this relationship
when considering total carbon stock because the
soil may contain over half of total ecosystem
carbon, thereby preventing a simple relationship
between the description of aboveground forest
characteristics and total carbon stock. Total carbon
stock in a muskeg meadow (ACPT 5), for example,
averages about 1.5 times that of unharvested high
volume old-growth forest (ACPT 18). However,
we did find a relationship between aboveground
and soil carbon stocks, one largely defined by the
following ecosystem types: muskeg, forest, and
alpine meadow/rock and ice (see Appendix 6 at
<http://www.springerlink.com>).
Total carbon in the Tongass (soil, aboveground
living biomass, and roots and dead woody debris)
was estimated to be 2.8 ± 0.5 Pg (95% confidence
interval [CI]) (Table 1). In all, 42% of the vari-
ability is the uncertainty in aboveground carbon
stock estimates, 6% is from uncertainty in root
carbon (root-to-shoot ratios), and 52% is from
uncertainty in soil carbon. Assumptions about the
allometric biomass equation used for willow and
birch, the exclusion of an outlying CWD data point,
and estimation of CIs for carbon stocks in polygons
lacking sufficient data have insignificant influence
on total carbon or the CI (Table 1). Trees outside
the size range of the allometric models account for
19% of the total carbon estimate. Three-dimen-
sional surface plots of the allometric equations
maintained consistent shape outside the dbh range
for which the equations were developed.
The carbon stock in the Tongass forest and soils
(2.8 Pg) comprises 7.7% of the carbon in the forests
and soils of the conterminous United States
(36.7 Pg) (Turner and others 1995) and 0.25% of
the carbon in the Earth’s forest vegetation and soils
(1,146 Pg) (Dixon and others 1994).
In all, 66% of the total carbon in the Tongass is
in the soils, 30% is in aboveground biomass (15%
in live trees, 6% in seedlings and saplings, 3% in
standing dead wood, 6% in CWD, less than 1% in
SWD, and 1% in understory vegetation), and 4% is
in roots. Less than 1% of the total carbon estimates
Figure 5. Aboveground carbon
stock by Aboveground Carbon
Polygon Type (ACPT) number,
ranked by aboveground carbon
stock. Carbon stocks for all
ACPTs (polygon types with nless
than 5 are omitted) are shown in
gray. Asterisks identify the 10
ACPTs that account for 86% of
total carbon in the Tongass (95%
CI).
Carbon Sequestration in Southeast Alaska Forest 1059
were influenced by the assumptions involved in
our calculation of aboveground carbon stocks (for
example, selection of allometric equations and
application of these equations beyond their speci-
fied range). Uncertainty in the density and distri-
bution of understory vegetation did not affect the
analysis. Twenty-two percent of total carbon in the
Tongass is in the soils of polygons where soil types
have not been mapped. Comparison of the results
from application of soil carbon density estimates
from Alexander and others (total soil car-
bon = 1.9 Pg) with total soil carbon given by cal-
culations using FIA Grid Inventory soil pit data
(total soil carbon = 0.49 Pg) suggests that more
than 70% of soil carbon is not reported in the FIA
data.
We produced several net carbon flux projections
for each management regime to capture carbon
dynamics associated with the following factors:
variations in the residence time of carbon in slash,
long-term forest products, and short-term forest
products; and effects of the carbon accretion model
(polynomial or asymptotic) (Figure 4). The annual
rate of net carbon flux is the first derivative of the
aggregate net carbon flux presented in Figure 4.
Doubling or halving the time periods for 90%
carbon emission from the saw timber, pulp prod-
ucts, and slash pools alters the shape of projected
net carbon flux curves but causes less than 0.6%
change in average annual net carbon flux for all
modeled management regimes.
The average annual net carbon flux from the
Tongass during the period 1900–54 was 60,000
Mg C/y, and the average annual net flux from the
Tongass for the subsequent 41-year period was
307,000 Mg C/y. Estimates of future net carbon
fluxes are presented in Table 2; upper- and lower-
bound estimates were calculated using the results
of the sensitivity analyses.
Our best estimate of the net annual economic
value of carbon sequestration that would result
from ceasing all harvesting in the Tongass is $4 to
$7 million/y for the 100-year period 1995–2095
and $3 to $6 million/y for the 200-year period
1995–2195 (Table 3). Our best estimate of the net
annual economic value of carbon emission result-
ing from increased harvesting of administratively
available forested lands is )$3 million/y for the
100-year period 1995–2095 and )$2 to )$4
million/y for the 200-year period 1995–2195.
D
ISCUSSION
Using GIS data in combination with FIA data
proved to be an effective and robust approach to
estimating carbon stocks and modeling the effects
of different management regimes on future net
carbon flux. New spatially explicit data could be
integrated into our existing models, enabling
application of the models to other areas and
refinement of net carbon flux estimates if future
GIS data collection is carried out with this
application in mind.
A lack of data on tree size and density necessi-
tated the use of timber volume classes in mapping
carbon stocks. Although tree size and density data
are preferable, timber volume is tightly correlated
with carbon stocks (Hamburg and others 1997),
and low variances among the 10 most important
ACPTs suggest the robustness of using existing
volume data to map carbon stocks.
The range in estimates of net carbon flux from
ceasing all timber harvesting may overestimate the
uncertainty in this projection. We aggregated
uncertainties of carbon stocks, assumptions about
aboveground carbon stocks postharvest, and the
carbon accretion model that we used; yet it
is highly likely that these uncertainties are
independent, and thus not additive.
The uncertainty in net flux estimates resulted
largely from selection of the biomass accretion
model, asymptotic or polynomial (Figure 2). The
rapidity with which carbon accretion progresses to
equilibrium in the asymptotic model may be
unrealistic, but the polynomial model’s prediction
of carbon stocks greater than those found in old-
growth stands may also be unrealistic. Unfortu-
nately, the limited availability of chronosequence
data leaves a gap in our understanding of carbon
accretion during the transition period from early
secondary growth to old growth. Furthermore,
calculation of carbon stocks for old-growth stands
from area-weighted averages of old-growth poly-
gon types is not analogous to the FIA permanent
plot data used for young stands and may confound
our accretion models. Data from FIA permanent
plots in old-growth forest could be used to test both
our assumption of steady-state carbon stocks and
250 Mg C/ha in aboveground live biomass in
old-growth forest. The actual pattern of carbon
accretion probably lies somewhere between the
polynomial and asymptotic models, but we have
insufficient data to craft a more realistic model
(Janisch and Harmon 2002).
Our use of the area-weighted average above-
ground carbon stocks of all old-growth commercial
forest types in creating the biomass accretion mod-
els could introduce bias if remaining old-growth
forests are lower in biomass than the old-growth
forests already harvested. Failure to area-weight
1060 W. W. Leighty and others
Table 2. Average Annual Net Carbon (C) Flux Projections from the Tongass
Average Annual Net C Flux from Tongass National
Forest (000s Mg C/y)
Average Annual Net C Flux from Tongass National
Forest to the Atmosphere (000s Mg C/y)
Aboveground Standing
Biomass C after
Harvest = 86 Mg C/ha
Aboveground Standing
Biomass C after
Harvest = 0 Mg C/ha
Aboveground Standing
Biomass C after
Harvest = 86 Mg C/ha
Aboveground Standing
Biomass C after
Harvest = 0 Mg C/ha
Management Regime
C accretion
Model
Upper
Bound
Best
Estimate
Lower
Bound
Upper
Bound
Best
Estimate
Lower
Bound
Upper
Bound
Best
Estimate
Lower
Bound
Upper
Bound
Best
Estimate
Lower
Bound
1900–95
Historic management Polynomial accretion 180 160 67 180 160 67 140 130 54 140 130 54
Asymptotic accretion 200 180 74 180 160 67 170 150 63 140 130 54
1995–2095
Cessation of all harvesting Polynomial accretion )330 )270 )210 )330 )270 )210 )330 )270 )210 )280 )210 )180
Asymptotic accretion )270 )200 )120 )210 )150 )91 )270 )200 )120 )210 )150 )91
All forested lands harvested Polynomial accretion 840 440 71 1700 1200 880 470 200 )65 1100 790 520
on a 100-y rotation Asymptotic accretion 1300 780 390 2300 1800 1300 530 310 130 1700 1300 910
All forested lands harvested Polynomial accretion )0.59 )110 )230 410 300 180 )160 )200 )280 140 95 12
on a 200-y rotation Asymptotic accretion 640 410 210 1200 910 650 280 170 84 920 710 480
Administratively available Polynomial accretion 160 5.4 )140 400 250 97 100 )15 )160 280 160 17
forested lands harvested on
a 100-y rotation
Asymptotic accretion 280 120 )12 600 420 230 91 0.43 )69 480 330 150
Administratively available Polynomial accretion )40 )100 )180 81 17 )58 )39 )87 )170 51 1.7 )86
forested lands harvested
on a 200-y rotation
Asymptotic accretion 150 57 )16 330 230 140 55 )1.5 )43 300 200 73
1995–2195
Cessation of all harvesting Polynomial accretion )190 )160 )130 )190 )160 )130 )190 )160 )130 )170 )130 )110
Asymptotic accretion )140 )100 )58 )100 )70 )44 )140 )100 )58 )100 )70 )44
All forested lands harvested
on a 100-y rotation
Polynomial accretion 740 430 150 1300 1000 720 770 450 160 1300 1000 740
Asymptotic accretion 1100 680 360 1600 1300 920 1100 690 360 1700 1300 930
All forested lands harvested
on a 200-y rotation
Polynomial accretion 120 )39 )180 610 450 300 42 )84 )210 470 350 220
Asymptotic accretion 850 540 280 1400 1100 770 670 420 210 1300 960 690
Administratively available Polynomial accretion 160 57 )47 330 230 120 190 83 )37 350 220 100
forested lands harvested
on a 100-y rotation
Asymptotic accretion 250 130 27 440 310 190 260 130 28 470 340 200
Administratively available Polynomial accretion )30 )110 )180 120 40 )33 )29 )97 )180 100 32 )48
forested lands harvested
on a 200-y rotation
Asymptotic accretion 190 76 )17 360 250 150 140 47 )30 350 230 110
Annual net carbon flux was projected for total carbon leaving (or entering) the Tongass and for the portion of this carbon exchanged with the atmosphere. To quantify the influence of two important uncertainties, carbon accretion rate and
carbon stock in aboveground standing biomass left after harvesting, net carbon flux was modeled for each management regime with both polynomial and asymptotic carbon accretion curves and with carbon in aboveground standing
biomass after harvesting equal to zero or 86 Mg C/ha. Our best estimate of each net flux was calculated with reasonable judgments for other required assumptions; the upper and lower bounds were calculated with the combination of
assumptions that yielded the highest and lowest possible net flux estimates.
Carbon Sequestration in Southeast Alaska Forest 1061
these mean values, however, could give too much
importance to the rare forest conditions, which
have relatively few representative FIA sample plots.
Net carbon flux projections for the 200-year
rotation scenarios are more strongly influenced by
selection of the carbon accretion model than are
the 100-year rotation scenarios because the
200-year rotations allow enough time for second-
ary growth to reach the peak carbon stocks pre-
dicted by the polynomial model. Net carbon flux
projections for management regimes involving
100-year rotations are less sensitive to the selection
of carbon accretion curve because forested lands
are reharvested before there is a significant differ-
ence in the trajectories of the two models. Reso-
lution of the uncertainty in carbon accretion rates
is imperative for informing forest management
policy directed at carbon sequestration.
The distribution of carbon among soils (66%),
aboveground living and dead biomass (30%), and
belowground living biomass (4%) is consistent
with carbon inventories completed in other eco-
systems (Turner and others 1995). The large pro-
portion of the carbon stocks found in soil is due to
large areas of muskeg and deep organic soils in
southeast Alaska and is consistent with the average
for other temperate forests (Prentice and others
2001). Our approach to estimating soil carbon re-
sulted in conservative estimates of the total carbon
stock in this pool. Mapping conventions may have
underestimated the depth of hemist soils in the
Tongass by classifying them as saprists (none of
which are deep), which would cause underesti-
mation of carbon stocks (D’Amore and Lynn 2002).
The large discrepancy in results from our two
methods of estimating soil carbon stocks suggests
severe underestimation when FIA data are used.
Consequently, we did not combine our estimates or
use them as separate lines of evidence in our
uncertainty analysis.
Uncertainty in the soil carbon stock, which rep-
resents about half of the uncertainty in total carbon
stock estimates, was not incorporated into our
estimates of net carbon flux because we assumed
equilibrium in soil carbon stocks. Although forest
harvesting has little effect on soil carbon on aver-
age, specific harvesting techniques can cause in-
creases or decreases in soil carbon (Johnson and
Curtis 2001). There is insufficient information,
however, on the effects of harvesting in south-
eastern Alaska to include soil carbon in our net flux
models. Carbon flux from soils could represent a
significant addition to the net carbon flux associ-
ated with harvesting in southeastern Alaska, but
the assumption of soil equilibrium is necessary
until more data are available.
In defining our ‘‘best estimates’’ of net carbon
flux for the management regimes modeled, we
made the following assumptions: zero carbon in
standing aboveground biomass after clear-cutting;
13% reduction of CER allocations for carbon
sequestration associated with cessation of harvest-
ing as a result of reduced carbon storage in long-
term forest products; and the 200-year rotation
represents the baseline case upon which CER
allocation is based (current forest management
equates to a 180-year rotation). These assumptions
significantly reduce the range in our net flux esti-
mates, but some uncertainties (for example, carbon
accretion model) persist.
Table 3. Average Annual Economic Values for Net Carbon Flux ($ million/y) from the Tongass to the
atmosphere
Secondary Growth Curve
Polynomial Accretion Asymptotic Accretion
Management Regime Modeled 1995–2095 1995–2195 1995–2095 1995–2195
Cessation of all harvesting 3.7 2.2 2.5 1.2
100-y rotation (all forested lands) )16 )21 )26 )26
200-y rotation (all forested lands) )1.9 )6.9 )14 )19
100-y rotation (admin. avail. forested lands) )3.2 )4.5 )6.6 )6.8
200-y rotation (admin. avail. forested lands) )0.03 )0.63 )4.0 )4.7
Maximum range of net annual carbon value from ceasing harvest 3.7–20 2.9–23 6.6–29 5.9–27
Best estimate of net annual carbon value from ceasing harvest 3.7 2.9 6.6 5.9
Average annual economic value of net carbon flux for each management regime modeled was calculated using our net carbon flux estimates and a value of $20 Mg
)1
C. The
maximum range of net annual carbon value from ceasing harvest is the difference between ceasing harvest and the alternative management regime with the most carbon
emission (100-year rotation of all forested lands). The best estimates of net annual economic value are the difference between ceasing harvest and 200-year rotation of
administratively available forest lands (a close approximation of ‘‘business as usual’’). These estimates assume zero carbon in standing aboveground biomass after harvesting
and reduction of Certified Emission Reduction Credits (CERs) by 13% to account for reduced carbon storage in long-term forest products.
1062 W. W. Leighty and others
Net carbon flux into or out of the Tongass is not
large enough to significantly impact the US carbon
budget. The US Environmental Protection Agency’s
(EPA) 2003 inventory of GHG emissions and sinks
estimated that net carbon flux from the forests of
the conterminous United States amounted to
267 Tg C/y in 1995 (US Environmental Production
Agency 2003). Our estimates for the Tongass of
0.13–1.8 Tg/y are 0.04%–0.7% of the EPA’s
inventory. Similarly, the potential for carbon
sequestration due to management change in the
Tongass is significantly less than that for other
options for land-use change. Cessation of all har-
vesting of available lands in the Tongass (1.3 ·10
6
ha) results in annual sequestration of 0.04–0.33
Tg C/y, or 31 to 250 kg C ha
)1
y
)1
. By comparison,
the land enrolled in the Conservation Reserve
Program (CRP) in 1996 (16.2 ·10
6
ha) may
sequester as much as 12 Tg C/y (Barker and others
1995), which is three to 30 times the rate per unit
area in the Tongass. However, the economic cost of
carbon sequestration in the Tongass may be sig-
nificantly less than that for the CRP. Assuming that
the lost revenue from US Forest Service timber
sales is the cost of carbon sequestration in the
Tongass, for example, the cost of carbon seques-
tration in the Tongass would be about one-quarter
of the CRP cost (approximately $0.02/kg C versus
approximately $0.08/kg C).
Past harvesting caused the net loss of 1.3–3.6
Tg C from the Tongass from 1900 to 1954 and 5.1–
13.6 Tg from 1954 to 1995; these numbers include
emissions from harvesting and sequestration from
regrowth. For comparison, land use in the conter-
minous United States caused the loss of
27,000 ± 6000 Tg carbon from 1900 to 1945, but
the regrowth of northeastern forests resulted in a
net gain of 2000 ± 2000 Tg C from 1954 to 1995
(Houghton and others 1999).
The conversion of 6 ·10
6
ha of old-growth
forest to young plantations in forests of Wash-
ington and Oregon is similar to the logging history
of the Tongass, and resulted in the loss of 1500–
1800 Tg C from aboveground and soil carbon
pools (Harmon and others 1990). Harvesting in
the Tongass has caused the loss, from above-
ground carbon pools only and net of subsequent
regrowth, of 13%–29% (6.4–17.2 Tg C on 0.2 ·
10
6
ha) of the carbon per hectare released from
the forests of Washington and Oregon. Harmon
and others use of Covington’s model of decline in
O horizon soil carbon after harvesting may have
led to a significant overestimate of the loss of soil
carbon (Yanai and others 2003). Our estimates of
net carbon flux from aboveground biomass (150–
210 Mg C/ha) are similar to those of Harmon and
others (187 Mg C/ha).
The economic value of carbon sequestration
associated with the cessation of harvesting in the
Tongass may be significant relative to the value of
the timber harvested. Our best estimates of the net
annual economic value of carbon sequestration
resulting from cessation of all harvesting in the
Tongass ($3 to $7 million/y) are of similar magni-
tude to the annual revenue from timber sales in the
Tongass ($6.5 million/y) (USDA Forest Service
2001). Potential cobenefits of harvesting timber
and of ceasing harvest (for example, fisheries,
tourism, timber processing) could influence the
total net annual economic value for each
management regime.
Some investigators have suggested that carbon
sequestration from land-use change may not mit-
igate climate change as effectively as the reduction
of GHG emissions from fossil fuel use, citing the
possibility for leakage (that is, emissions associated
with production may be displaced to another
location). Reduced harvesting in the Tongass may
require increased harvesting elsewhere to keep
product supply constant. Consequently, estimates
of the monetary value to Tongass managers for
carbon sequestration may not reflect the net social
benefit nor the benefit to the USDA Forest Service
if another national forest increases its harvesting,
buying CERs to do so, to keep the total product
stream from national forest lands constant.
The net economic value of carbon sequestration
associated with the elimination of harvesting in the
Tongass clearly depends on the value of CERs. This
value was assumed to be $20 Mg
)1
C, but estimates
of the value of CERs in a regulated marketplace
range from $5 to $125 Mg
)1
C (Weyant 2000).
Deviation in the value of CERs from $20 Mg
)1
C
was not included in the estimated range of net
economic value from carbon sequestration in the
Tongass because the range scales linearly.
Some additional factors omitted from our anal-
yses deserve mention. First, increasing atmospheric
concentration of carbon dioxide and changing re-
gional climates may alter some characteristics of
the Tongass, including carbon stock and flux.
However, the magnitude of changes in carbon
stock caused by climate change is small compared
to changes caused by land use (Caspersen and
others 2000; Houghton and others 1999). Second,
the assumption of steady-state carbon stocks in old-
growth forests is ubiquitous, despite a dearth of
data available to either confirm or disprove it, for
Alaska or elsewhere. Third, young forests generally
have lower levels of defect from decay than old-
Carbon Sequestration in Southeast Alaska Forest 1063
growth forests. Consequently, the proportions of
harvested material used in forest product streams
may change with conversion of forested lands in
the Tongass from old-growth forest to managed
younger stands, with implications for the question
of whether harvesting less area more intensely re-
sults in greater carbon storage than harvesting
more area less intensely. Fourth, the possibilities
for improving efficiency in timber harvesting (Fa-
hey 1983) were not included in our models because
they are highly dependent on a large number of
economic variables that are beyond the scope of
this research. Finally, changes in species composi-
tion, caused by management or climate change,
could influence carbon flux due to associated shifts
in the relative importance of white and brown rots
in wood decay (Kimmey 1956).
The Tongass must be included in accurate na-
tional carbon budgets. Furthermore, management
of the Tongass for carbon sequestration may be of
equivalent economic value to timber harvesting.
Valuation of potential carbon sequestration in the
Tongass from ceasing all harvesting may be ampli-
fied by indirect benefits of eliminating harvesting,
such as maintenance of the southeast Alaska fish-
eries and tourism industries and reduced expenses
for the Tongass timber program. Complete valua-
tion of timber harvesting may be influenced by
cobenefits as well. The emerging economic value of
carbon sequestration requires consideration of net
carbon flux in the development of future Tongass
management plans.
ACKNOWLEDGEMENTS
W.W.L. received financial support from the Royce
Fellowship program at Brown University. We
thank Mike McClellan, Frances Biles, and Dave
D’Amore at the US Forest Service Juneau Forestry
Sciences Laboratory for invaluable help with data
collection and manipulation. Mark Harmon, Linda
Heath, and two anonymous reviewers provided
insightful comments on an earlier version of the
manuscript.
REFERENCES
Alemdag IS. 1984. Total tree and merchantable stem biomass
equations for Ontario hardwoods. Informational Report PI-X-
46. Canadian Forestry Service. Canada Chalk River, Ontario:
Petawawa National Forestry Institute, Canadian Forestry
Service, Agriculture.
Alexander EB, Kissinger E, Huecker RH, Cullen P. 1989. Soils of
southeast Alaska as sinks for organic carbon fixed from
atmospheric carbon-dioxide. p. 203-210 In E.B. Alexander
(ed.) Proc. of Watershed ’89: A conference on the Stewardship
of Soil, Air, and Water Resources. 21-23 Mar. 1989. Juneau,
Alaska. USDA Forest Service General Technical Report No.
R10-MB-77. Juneau, Alaska: USDA Forest Service.
Barker JR, Baumgardner GA, Turner DP, Lee JJ. 1995. Potential
carbon benefits of the Conservation Reserve Program in the
United States. J Biogeogr 22:743–51.
Brown JK. 1974. Handbook for inventorying downed woody
material. General Technical Report INT-GTR-16. US Depart-
ment of Agriculture. Intermountain Forest and Range
Experiment Station Fort Collins, Colorado.
Cairns MA, Brown S, Helmer EH, Baumgardner GA. 1997. Root
biomass allocation in the world’s upland forests. Oecologia
111:1–11.
Caspersen JP, Pacala SW, Jenkins JC, Hurtt GC, Moorcraft PR,
Birdsey RA. 2000. Contributions of land-use history to carbon
accumulation in U.S. forests. Science 290:1148–51.
D’Amore DV, Lynn WC. 2002. Classification of forested Histosols
in southeast Alaska. Soil Sci Soc Am J 66:554–62.
DeMars DJ. 2000. Stand-density study of spruce-hemlock stands
in southeastern Alaska. General Technical Report PNW-GTR-
496. US Department of Agriculture. Ogden, Utah: Pacific
Northwest Research Station.
Dixon RK, Brown S, Houghton RA, Solomon AM, Trexler MC,
Wisniewski J. 1994. Carbon pools and flux of global forest
ecosystems. Science 263:185.
Everest FH, Swanston DN, Shaw CG III, Smith WP, Julin KR,
Allen SD. 1997. Evaluation of the use of scientific information
in developing the 1997 forest plan for the Tongass National
Forest. General Technical Report PNW-GTR-415. US Depart-
ment of Agriculture Forest Service. Portland, Oregon: Pacific
Northwest Research Station.
Fahey TD. 1983. Product recovery from hemlock ‘‘pulpwood’’
from Alaska. Research Paper PNW-303. US Department of
Agriculture. Portland, Oregon: Pacific Northwest Forest and
Range Experiment Station.
Hamburg SP, Zamolodchikov DG, Korovin GN, Nefedjev VV,
Utkin AI, Gulbe JI, Gulbe TA. 1997. Estimating the carbon
content of Russian forests: a comparison of phytomass/vol-
ume and allometric projections. Mitigation and Adaptation
Strategies for Global Change Netherlands: Springer 2:247–65.
Harmon ME, Ferrell WK, Franklin JF. 1990. Effects on carbon
storage of conversion of old-growth forests to young forests.
Science 247:699–702.
Houghton RA, Hackler JL, Lawrence KT. 1999. The U.S. carbon
budget: contributions from land-use change. Science
285:574–7.
Janisch JE, Harmon ME. 2002. Successional changes in live and
dead wood carbon stores: implications for net ecosystem
productivity. Tree Physiol 22:77–89.
Johnson DW, Curtis PS. 2001. Effects of forest management on
soil C and N storage: meta-analysis. For Ecol Manage
140:227–38.
Kimmey JW. 1956. Cull factors for Sitka spruce, western hem-
lock and western red cedar in southeast Alaska. Station Paper
No. 6. US Department of Agriculture. Juneau, Alaska: Alaska
Forest Research Center.
Krumlik, GJ. 1974. Biomass and nutrient distribution in two old
growth forest ecosystems in south coastal British Columbia.
Masters thesis, University of British Columbia, Vancouver,
B.C.
Nowacki G, Shephard M, Krosse P, Pawuk W, Fisher G, Baichtal
J, Brew D, and others. 2001. Ecological subsections of south-
east Alaska and neighboring areas of Canada. Technical Pub-
1064 W. W. Leighty and others
lication R10-TP-75. US Department of Agriculture Forest
Service. Alaska Region, Juneau, Alaska: USDA Forest Service.
Prentice IC, Farquhar GD, Fasham MJR, Goulden ML, Heimann
M, Jaramillo VJ, Kheshgi HS, and others. 2001. The carbon
cycle and atmospheric carbon dioxide. Houghton JT, Ding Y,
Griggs DJ, Noguer M, van der Linden PJ, Dai X, Maskell K,
and others. Eds. Climate change 2001: the scientific
basis.Cambridge University Press, Cambridge (UK)183238.
Sampson RN, Hair D, Eds. 1996. Forest management opportu-
nities for mitigating carbon emissions. Forest and global
change; vol 2. Washington, D.C.: American Forests.
Santantonio D, Hermann RK, Overton WS. 1977. Root biomass
studies in forest ecosystems. Pedobiologia 17:1–31.
Shaw, DL. 1977. Biomas equations for Douglas-fir, western
redcedar, and red alder in Washington and Oregon. Tech. Rep.
044-3001/77/22. Centralia, WA: Weyerhaeuser Corp.
Singh T. 1983. Weight tables for important tree species in the
Northwest Territories. Report on. NOR FMN-27. Canadian
Forestry Service. Edmonton, Alberta: Northern Forest
Research Centre.
Skog KE, Nicholson GA. 1998. Carbon cycling through wood
products: the role of wood and paper products in carbon
sequestration. For Prod J 48(7):75–84.
Smithwick EAH, Harmon ME, Remillard SM, Acker SA, Franklin
JF. 2002. Potential upper bounds of carbon stores in forests of
the pacific northwest. Ecol Appl 12(5):1303–17.
Standish JT, Manning GH, Demaerschalk JP. 1983. Development
of biomass equations for British Columbia tree species.
Informational Report BC-X-264. Canadian Forestry Service.
Victoria, B.C.: Pacific Forest Research Centre.
Turner DP, Koerper GJ, Harmon ME, Lee JJ. 1995. A carbon
budget for forests of the conterminous United States. Ecol
Appl 5(2):421–36.
US Forest Products Laboratory. 1974. Wood handbook: wood as
an engineering material. Agricultural Handbook 72. US
Department of Agriculture. Washington, D.C.: US Govern-
ment Printing Office.
USDA Forest Service. 2005. Tongass National Forest facts.
Available online at: www.fs.fed.us/r10/tongass/.
USDA Forest Service. 2000. Tongass National Forest. Available
online at: Existing_Veg [vector digital data]. http://permanent.
access.gpo.gov/websites/fsfedus/www.fs.fed.us/r10/tongass/
gisinfo/page.
USDA Forest Service. 2001. Timber Sale Program Information
Reporting System (TSPIRS), fiscal year 1998 TSPIRS docu-
ments. Available online at: www.fs.fed.us/land/fm/tspirs/
1998tspirs/.
USDA Forest Service. 1995a. Timber supply and demand. Alaska
National Interest Lands Conservation Act. Section 706(a).
Report to Congress no. 15. Alaska National Interest Lands
Conservation Act, Sect. 706(a) Report to Congress. Alaska
Region, Juneau, Alaska: US Forest Service.
USDA Forest Service. 1995b. Field procedures for the southeast
Alaska inventory 1995. Juneau, Alaska: Pacific Northwest
Research Station Forestry Sciences Laboratory and Region 10.
USDA Soil Conservation Service. 1992a. Classification and cor-
relation of the soils of the Stikine area, Alaska. Report no.
SSSA-645. US Department of Agriculture. Petersburg Area,
Anchorage, Alaska: USDA Forest Service.
USDA Soil Conservation Service. 1992b. Classification and cor-
relation of the soils of the Chatham area, Alaska. Report no.
SSSA-646. US Department of Agriculture. Sitka Area,
Anchorage, Alaska: USDA Forest Service.
USDA Soil Conservation Service. 1994. Classification and cor-
relation of the soils of the Ketchikan area, Alaska. Report no.
SSSA-644. US Department of Agriculture. Ketchikan Area,
Anchorage, Alaska: USDA Forest Service.
US Environmental Protection Agency. 2003. Inventory of U.S.
greenhouse gas emissions and sinks: 1990–2001. Washington
(DC): US Environmental Protection Agency.
Waddel KL. Public Communication 2001. An application of line
interest sampling to estimate attributes of coarse woody debris
in resource inventories, USDA Forest Service Pacific North-
west research Station, Forest Sciences Laboratory, Portland,
Oregon.
Warren DD. 1999. Production, prices, employment, and trade
in northwest forest industries, fourth quarter 1997.
Resource Bulletin PNW-RB-230. US Department of Agri-
culture. Portland, Oregon: Pacific Northwest Research
Station.
Watson RT, Noble IR, Bolin B, Ravindranath NH, Verardo DJ,
Dokken DJ, Eds. 2000. Land use, land-use change, and for-
estry. Cambridge (UK): Cambridge University Press.
Weyant JP. 2000. An introduction to the economics of climate
change policy. Arlington (VA): Pew Center on Global Climate
Change, 56.
Yanai RD, Currie WS, Goodale CL. 2003. Soil carbon dynamics
following forest harvest: an ecosystem paradigm reviewed.
Ecosystems 6:197–212.
Yarie J, Mead BR. 1988. Twig and foliar biomass estimation
equations for major plant species in the Tanana River basin of
interior Alaska. Research Paper PNW-RP-401. US Department
of Agriculture. Portland, Oregon: Pacific Northwest Research
Station.
Carbon Sequestration in Southeast Alaska Forest 1065
... Carbon (C) stocks have been quantified previously on the TNF [12] and recognized as nationally significant by USDA Forest Service researchers [13][14][15] and in congressional policy reviews [16]. However, the USDA Forest Service has undervalued the C stock importance of the TNF by routinely dismissing stock change from logging as inconsequential to total US greenhouse gases (GHGs) [10,17]. ...
... Our analysis is key to shedding light on the importance of IRA protections and policy options for both old growth and young-growth forests. Given the national significance of C stocks on the TNF [12], managing forests to maximize C stock potential would demonstrate the US has made a forest-based nationally determined contribution (NDC) to the Paris Climate Agreement. Article 5.1 of the agreement recognizes the need for countries to take specific actions that conserve and enhance nature-based solutions as C sinks and reservoirs [19]. ...
... Values in parentheses indicate ranges (lower and upper bounds). Biomass was scaled [25] to determine lower and upper bounds using the range of ratios between the live trees measured by Forest Inventory Analysis (FIA) plot data and the other C pools (excluding soils) [12]. Soil was not scaled (see [28]), hence the lack of ranges. ...
Article
Full-text available
The 6.7 M ha Tongass National Forest in southeast Alaska, USA, supports a world-class salmon fishery, is one of the world’s most intact temperate rainforests, and is recognized for exceptional levels of carbon stored in woody biomass. We quantified biomass and soil organic carbon (C) by land use designation, Inventoried Roadless Areas (IRAs), young and productive old-growth forests (POGs), and 77 priority watersheds. We used published timber harvest volumes (roundwood) to estimate C stock change across five time periods from early historical (1909–1951) through future (2022–2100). Total soil organic and woody biomass C in the Tongass was 2.7 Pg, representing ~20% of the total forest C stock in the entire national forest system, the equivalent of 1.5 times the 2019 US greenhouse gas emissions. IRAs account for just over half the C, with 48% stored in POGs. Nearly 15% of all C is within T77 watersheds, >80% of which overlaps with IRAs, with half of that overlapping with POGs. Young growth accounted for only ~5% of the total C stock. Nearly two centuries of historical and projected logging would release an estimated 69.5 Mt CO2e, equivalent to the cumulative emissions of ~15 million vehicles. Previously logged forests within IRAs should be allowed to recover carbon stock via proforestation. Tongass old growth, IRAs, and priority watersheds deserve stepped-up protection as natural climate solutions.
... Temperate rainforests have high carbon densities compared to other forests (Keith et al., 2009). Southeast Alaska contains an estimated 2.8 ± 0.5 Pg C (Leighty et al., 2006), with 1.21-1.52 Pg C in aboveground biomass (Buma & Thompson, 2019) and approximately 1.8 Pg SOC in an area of approximately 70,586 km 2 (McNicol et al., 2019). ...
... Pg C in aboveground biomass (Buma & Thompson, 2019) and approximately 1.8 Pg SOC in an area of approximately 70,586 km 2 (McNicol et al., 2019). The montane forests of SE Alaska are mostly undisturbed by human activities and are carbon-dense, estimated to contain 8% of all terrestrial biosphere carbon in the contiguous United States and 0.25% globally (Leighty et al., 2006). Parts of the region have been logged, resulting in a C loss of 60,000-307,000 tC yr −1 (0.85-4.35 tC km −2 yr −1 ) from 1900 to 1995 (Leighty et al., 2006), but a majority of land contains old-growth forests (Berg et al., 2014). ...
... The montane forests of SE Alaska are mostly undisturbed by human activities and are carbon-dense, estimated to contain 8% of all terrestrial biosphere carbon in the contiguous United States and 0.25% globally (Leighty et al., 2006). Parts of the region have been logged, resulting in a C loss of 60,000-307,000 tC yr −1 (0.85-4.35 tC km −2 yr −1 ) from 1900 to 1995 (Leighty et al., 2006), but a majority of land contains old-growth forests (Berg et al., 2014). Carbon flux from the atmosphere to living biomass is approximately 85 tC km −2 yr −1 , which should be roughly comparable in magnitude to the flux of carbon respired to the atmosphere from soil, assuming equilibrium. ...
Article
Full-text available
Landslides, a forest disturbance, mobilize carbon (C) sequestered in vegetation and soils. Mobilized C is deposited either onto hillslopes or into the water, sequestering C from and releasing C to the atmosphere at different time scales. The C‐dense old‐growth temperate forests of SE Alaska are a unique location to quantify C mobilization rate by frequent landslides that often evolve into saturated moving masses known as debris flows. In this study, the amount of C mobilized by debris flows over historic time scales was estimated by combining a landslide inventory with maps of modeled biomass and soil carbon. We analyzed SE Alaskan landslides over a 55‐year period where a total of 4.69 ± 0.21 MtC was mobilized, an average rate of 2.5 tC km⁻² yr⁻¹. A single event in August 2015 mobilized 57,651 ± 3,266 tC, an average of 63 tC km⁻². Depositional fate was inferred using two methods, a standard stream intersection analysis and a second novel approach using simulated debris flow deposition modeling calibrated to the study area. Approximately 60% of debris flow deposits intersected the stream network (9% into mainstem channels, 91% into small tributaries), consistent with long‐term modeled connectivity, suggesting that debris flows are likely to contribute to globally significant amounts of C buried in local fjord sediments. Our results are consistent with an emerging consensus that landslide disturbances that mobilize organic carbon may play an important role in the global carbon cycle over geologic time, with coastal temperate forests being hotspots of potential carbon sequestration.
... The type or description of impact by a disturbance on forest attributes was categorized as positive, negative, direct, or indirect (Fig. 8). Negative impacts were dominant, constituting 20% of the publications reviewed (e.g., Leighty et al., 2006;Coomes and Allen, 2007;Joyce et al., 1995). The analyses of positive impacts constituted only 10% (see Shifley et al., 2008;Gormley et al., 2012), although several studies also reported on both negative and positive impacts. ...
... • Undertake management practices that sustain or increase biodiversity and reduce disturbances in forest ecosystems (Alfaro-Sánchez et al., 2018;Barrette et al., 2014;Spies et al., 2017), including restricting human-induced disturbances in forest ecosystems (Garet et al., 2012;Leighty et al., 2006;Zhang et al., 2014) and second-growth forests (Adum et al., 2013); • Maintain connectivity and relationships between different elements in the forest to sustain the integrity of the forest ecosystem and prevent habitat fragmentation (Akresh and King, 2016;Bridger et al., 2016); • Identify and expand the variables considered important in forest management and landscape planning (Battles et al., 2001;Febbraro et al., 2015;Wilson et al., 2014;and Josefsson et al., 2009;Tata, 2021); • Use scientific approaches, measurements, and methodologies that consider historical information (James et al., 2017) to inform management practices (Carvajal et al., 2018;Vicente-Serrano et al., 2014); • Forest management practices must be based on a sound understanding of how forest ecosystems respond to various disturbances to enable forests to adapt to the changes (Vilà-Cabrera et al., 2011), including evaluating management practices to reflect reality (Jacqmain et al., 1999) and strengthening assessments regime of mining projects to systematically consider the cumulative impacts of forest loss (Siqueira-Gay and Sánchez, 2020); • Provide adequate infrastructure and resources to address threats to the forest from disturbances (Pinho et al., 2018;Reilly et al., 2018;Vadjunec and Rocheleau, 2009). • Provide adequate financial and human resources and management tools for proper enforcement and implementation of conservation goals (Remm et al., 2013;Vanonckelen and Rompaey, 2015). ...
Article
Full-text available
This paper reviews trends in the academic literature on cumulative effects assessment (CEA) of disturbance on forest ecosystems to advance research in the broader context of impact assessments. Disturbance is any distinct spatiotemporal event that disrupts the structure and composition of an ecosystem affecting resource availability. We developed a Python package to automate search term selection, write search strategies, reduce bias and improve the efficient and effective selection of articles from academic databases and grey literature. We identified 148 peer-reviewed literature published between 1986 and 2022 and conducted an inductive and deductive thematic analysis of the results. Our findings revealed that CEA studies are concentrated in the global north, with most publications from authors affiliated with government agencies in the USA and Canada. Methodological and analytical approaches are less interdisciplinary but mainly quantitative and expert-driven, involving modeling the impacts of disturbances on biophysical valued components. Furthermore, the assessment of socioeconomic valued components, including the effects of disturbance on Indigenous wellbeing connected to forests, has received less attention. Even though there is a high preference for regional assessment, challenges with data access, quality, and analysis, especially baseline data over long periods, are hampering effective CEA. Few articles examined CEA – policy/management nexus. Of the few studies, challenges such as the inadequate implementation of CEA mitigation strategies due to policy drawbacks and resource constraints, the high cost of monitoring multiple indicators, and poor connections between scenarios/modeling and management actions were paramount. Future CEA research is needed to broaden our understanding of how multiple disturbance affects forests in the global south and coupled social and ecological systems and their implications for sustainable forest management.
... There is more old-growth than young forest area in Alaska (Pan et al., 2011), and the most effective measure to protect existing carbon stocks is to protect and maintain existing old growth stands. The total carbon stock in the forest and soils of the Tongass National Forest in southeast Alaska alone is 8% of that of the forests in the conterminous United States (Leighty et al., 2006). Protecting forests with old trees will also confer co-benefits to wildlife, biodiversity, and other ecosystem services including the provision of microclimate buffering under future extreme climate conditions and various cultural ecosystem services (Sutherland et al., 2016). ...
... Average carbon storage in live vegetation and soils (2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014) was 131,675 MMT CO 2 Eq. on Alaska's federal lands, with 92 percent stored in soils and eight percent stored in live vegetation (Merrill et al., 2018). In the Tongass National Forest, 26 percent is stored in live vegetation (Leighty et al., 2006). Maps of total biomass carbon, which integrate above and below ground sources, indicate that nearly all of Alaska is above the global median (Figures 4, 5). ...
Article
Full-text available
Alaska is globally significant for its large tracts of intact habitats, which support complete wildlife assemblages and many of the world’s healthiest wild fisheries, while also storing significant amounts of carbon. Alaska has 1/3 of United States federal lands, the bulk of the United States’ intact and wild lands, and over half of the country’s total terrestrial ecosystem carbon on federal lands. Managing Alaska’s public lands for climate and biodiversity conservation purposes over the next 30–50 years would provide meaningful and irreplaceable climate benefits for the United States and globe. Doing so via a co-management approach with Alaska’s 229 federally recognized tribes is likely not only to be more effective but also more socially just. This paper lays out the scientific case for managing Alaska’s public lands for climate stabilization and resilience and addresses three primary questions: Why is Alaska globally meaningful for biodiversity and climate stabilization? Why should Alaska be considered as a key element of a climate stabilization and biodiversity conservation strategy for the United States? What do we need to know to better understand the role of Alaska given future scenarios? We summarize evidence for the role Alaska’s lands play in climate stabilization, as well as what is known about the role of land management in influencing carbon storage and sequestration. Finally, we summarize priority research that is needed to improve understanding of how policy and management prescriptions are likely to influence the role Alaska plays in global climate stabilization and adaptation.
... Alaskan coastal forests store the largest amount of carbon per unit area in the world when soils are included in the total (Heath et al., 2011;McNicol et al., 2019), with the carbon stored in large-stature conifers exceeded by belowground carbon storage in soils (Leighty, Hamburg and Caouette, 2006). Alaska's coastal forests are regarded as a carbon reservoir, but active management occurs in specific management zones on both public and private lands. ...
Book
Full-text available
Many people worldwide lack adequate access to clean water to meet basic needs, and many important economic activities, such as energy production and agriculture, also require water. Climate change is likely to aggravate water stress. As temperatures rise, ecosystems and the human, plant, and animal communities that depend on them will need more water to maintain their health and to thrive. Forests and trees are integral to the global water cycle and therefore vital for water security – they regulate water quantity, quality, and timing and provide protective functions against (for example) soil and coastal erosion, flooding, and avalanches. Forested watersheds provide 75 percent of our freshwater, delivering water to over half the world’s population. The purpose of A Guide to Forest–Water Management is to improve the global information base on the protective functions of forests for soil and water. It reviews emerging techniques and methodologies, provides guidance and recommendations on how to manage forests for their water ecosystem services, and offers insights into the business and economic cases for managing forests for water ecosystem services. Intact native forests and well-managed planted forests can be a relatively cheap approach to water management while generating multiple co-benefits. Water security is a significant global challenge, but this paper argues that water-centered forests can provide nature-based solutions to ensuring global water resilience.
... The carbon pool of forest vegetation accounts for more than 86% of the global vegetation carbon pool [7], and is an important focus for studying CO 2 balance and exchange between vegetation and the atmosphere [8,9]. Investigating the changes in the forest vegetation carbon pool allows us not only to evaluate carbon sink capacity [10] but also to understand the dynamics of the forest ecosystem [11,12]. It is, therefore, important to study the changes in forest vegetation carbon storage within a particular time scale. ...
Article
Full-text available
According to the forest resources inventory data for different periods and the latest estimation parameters of forest carbon reserves in China, the carbon reserves and carbon density of forest biomass in the Tibet Autonomous Region from 1999 to 2019 were estimated using the IPCC international carbon reserves estimation model. The results showed that, during the past 20 years, the forest area, forest stock, and biomass carbon storage in Tibet have been steadily increasing, with an average annual increase of 1.85×10 4 hm 2 , 0.033×10 7 m 3 , and 0.22×10 7 t, respectively. Influenced by geographical conditions and the natural environment , the forest area and biomass carbon storage gradually increased from the northwest to the southeast, particularly in Linzhi and Changdu, where there are many primitive forests, which serve as important carbon sinks in Tibet. In terms of the composition of tree species, coniferous forests are dominant in Tibet, particularly those containing Abies fabri, Picea asperata, and Pinus densata, which comprise approximately 45% of the total forest area in Tibet. The ecological location of Tibet has resulted in the area being dominated by shelter forest, comprising 68.76% of the total area, 64.72% of the total forest stock, and 66.34% of the total biomass carbon reserves. The biomass carbon storage was observed to first increase and then decrease with increasing forest age, which is primarily caused by tree growth characteristics. In over-mature forests, trees' photosynthesis decreases along with their accumulation of organic matter, and the trees can die. In addition, this study also observed that the proportion of mature and over-mature forest in Tibet is excessively large, which is not conducive to the sustainable development of forestry in the region. This problem should be addressed in future management and utilization activities.
... The carbon pool of forest vegetation accounts for more than 86% of the global vegetation carbon pool [7], and is an important focus for studying CO 2 balance and exchange between vegetation and the atmosphere [8,9]. Investigating the changes in the forest vegetation carbon pool allows us not only to evaluate carbon sink capacity [10] but also to understand the dynamics of the forest ecosystem [11,12]. It is, therefore, important to study the changes in forest vegetation carbon storage within a particular time scale. ...
Article
Full-text available
According to the forest resources inventory data for different periods and the latest estimation parameters of forest carbon reserves in China, the carbon reserves and carbon density of forest biomass in the Tibet Autonomous Region from 1999 to 2019 were estimated using the IPCC international carbon reserves estimation model. The results showed that, during the past 20 years, the forest area, forest stock, and biomass carbon storage in Tibet have been steadily increasing, with an average annual increase of 1.85×10 ⁴ hm ² , 0.033×10 ⁷ m ³ , and 0.22×10 ⁷ t, respectively. Influenced by geographical conditions and the natural environment, the forest area and biomass carbon storage gradually increased from the northwest to the southeast, particularly in Linzhi and Changdu, where there are many primitive forests, which serve as important carbon sinks in Tibet. In terms of the composition of tree species, coniferous forests are dominant in Tibet, particularly those containing Abies fabri , Picea asperata , and Pinus densata , which comprise approximately 45% of the total forest area in Tibet. The ecological location of Tibet has resulted in the area being dominated by shelter forest, comprising 68.76% of the total area, 64.72% of the total forest stock, and 66.34% of the total biomass carbon reserves. The biomass carbon storage was observed to first increase and then decrease with increasing forest age, which is primarily caused by tree growth characteristics. In over-mature forests, trees’ photosynthesis decreases along with their accumulation of organic matter, and the trees can die. In addition, this study also observed that the proportion of mature and over-mature forest in Tibet is excessively large, which is not conducive to the sustainable development of forestry in the region. This problem should be addressed in future management and utilization activities.
... Upland forests predominate in the seasonal NPCTR, where soils tend to be better drained. Across the NPCTR, aboveground C storage is high compared with tropical and boreal forests (Keith et al. 2009), estimated from 325 Mg per ha (95% confidence interval [CI]= 50) at the northern end (Alaska; Leighty et al. 2006), and increasing southward to 455 Mg per ha (95% CI = 156; BC; Matsuzaki et al. 2013) and 685 Mg per ha (95% CI = 47; Oregon and Washington; Smithwick et al. 2002). Forests of the NPCTR have been a relatively stable, large C sink over the past several decades (Peng et al. 2014, Buma andBarrett 2015), but sensitivity to climatic and hydrologic forcing means that thresholds in C production and storage are likely to be crossed as the climate continues to change. ...
Article
Full-text available
How climate change may affect land-sea linkages along the Pacific Coast and the ecological consequences of these changes for marine food webs and ecosystem processes .
... The role of forests and forestry for sequestering carbon and reducing greenhouse gas emissions is essential [13]. The carbon sequestration capacity of forests is also linked to management regimes at single tree, forest stand, or landscape levels [5,[14][15][16][17]. ...
Article
Full-text available
Lithuanian forestry has long been shaped by the classical normal forest theory, aiming for even long-term flow of timber, and the aspiration to preserve domestic forest resources, leading to very conservative forest management. With radically changing forest management conditions, climate change mitigation efforts suggest increasing timber demands in the future. The main research question asked in this study addresses whether current forest management principles in Lithuania can secure non-decreasing long-term flow of timber and carbon accumulation. The development of national forest resources and forestry was simulated for the next century using the Kupolis decision support system and assuming that current forest management is continued under the condition of three scenarios, differing by climate change mitigation efforts. Potential development trends of key forest attributes were analysed and compared with projected carbon stock changes over time, incorporating major forest carbon pools—biomass, harvested wood products and emission savings due to energy and product substitution. The key finding was that the total carbon balance should remain positive in Lithuania during the next one hundred years; however, it might start to decrease after several decades, with steadily increasing harvesting and a reduced increase of forest productivity. Additionally, incorporating the harvested wood and CO2 emissions savings in carbon balance evaluations is essential.
Technical Report
Full-text available
The Tongass National Forest in southeast Alaska previously relied on timber volume, a measure related to timber production and economics, to provide information on forest structure, ecosystem diversity, and wildlife habitat. The Forest Service developed a measure (Size Density Model [SDM]) based on tree density and mean tree diameter as a more comprehensive assessment of these key characteristics. To fully incorporate the SDM into planning and management of wildlife habitat and populations, information was needed on the relationship of land cover classes described by the SDM and habitat for wildlife species of conservation concern.
Article
Full-text available
The Tongass National Forest is the largest remaining relatively unaltered coastal temperate rain forest in the World. The Forest consists of 16.9 million acres of land distributed across more than 22 000 islands and a narrow strip of mainland in southeast Alaska. The Forest contains abundant timber, wildlife, fisheries, mineral, and scenic resources. The authors participated as scientists on the Tongass Land Management Planning Team from 1995 to 1997. We joined the planning team as full members but maintained separate and distinct roles from National Forest System members. We were asked to assure that credible, value-neutral, scientific information was developed independently without reference to management decisions. We also displayed the likely levels of risk to resources and society associated with various management options. We examined how scientific information was used in making management decisions and evaluated whether the decisions were consistent with the available information. We developed and used a set of criteria to evaluate the way in which managers used scientific information in formulating decisions. This evaluation began while the final alternative was in the formative stages so that managers could alter their management approach, if they so desired, before the Forest plan was finalized. Many management decisions were altered during this 'adaptive decisionmaking process' in which changes were made concurrent with iterations of this paper. Our conclusion was that the final management decisions made in developing the 1997 Forest plan achieved a high degree of consistency with the available scientific information. This paper does not consider any information gathered after the signing of the record of decision on May 23, 1997, or deal with subsequent implementation of the 1997 Tongass Forest plan.
Article
Placing an upper bound to carbon (C) storage in forest ecosystems helps to constrain predictions on the amount of C that forest management strategies could sequester and the degree to which natural and anthropogenic disturbances change C storage. The potential, upper bound to C storage is difficult to approximate in the field because it requires studying old-growth forests, of which few remain. In this paper, we put an upper bound (or limit) on C storage in the Pacific Northwest (PNW) of the United States using field data from old-growth forests, which are near steady-state conditions. Specifically, the goals of this study were: (1) to approximate the upper bounds of C storage in the PNW by estimating total ecosystem carbon (TEC) stores of 43 old-growth forest stands in five distinct biogeoclimatic provinces and (2) to compare these TEC storage estimates with those from other biomes, globally. Finally, we suggest that the upper bounds of C storage in forests of the PNW are higher than current estimates of C stores, presumably due to a combination of natural and anthropogenic disturbances, which indicates a potentially substantial and economically significant role of C sequestration in the region. Results showed that coastal Oregon stands stored, on average, 1127 Mg C/ha, which was the highest for the study area, while stands in eastern Oregon stored the least, 195 Mg C/ha. In general, coastal Oregon stands stored 307 Mg C/ha more than coastal Washington stands. Similarly, the Oregon Cascades stands stored 75 Mg C/ha more, on average, than the Washington Cascades stands. A simple, area-weighted average TEC storage to I m soil depth (TEC,,,) for the PNW was 671 Mg C/ha. When soil was included only to 50 cm (TEC(50)), the area-weighted average was 640 Mg C/ha. Subtracting estimates of current forest C storage from the potential, upper bound of C storage in this study, a maximum of 338 Mg C/ha (TEC(100)) could be stored in PNW forests in addition to current stores.
Article
As increased attention is being paid to the role Russian forests play in the global carbon budget, the desirability of being able to accurately and easily estimate the carbon content of the Russian forests is clear. In Russia timber volume has been estimated regularly and systematically as part of the former Soviet Union's forest inventory system. To determine the accuracy of using a volumetric approach to determining forest carbon pools, we developed allometric equations for the dominant trees (5 taxa) of two regions (Vologda and Volgograd) of Russia. Using these allomctric equations and the phytomass/volume ratios previously developed to exploit the volume inventory data, we compared the forest carbon content of 51 forest stands of varying ages, composition and structure estimated using the two approaches. Carbon estimates for the Vologda region were on average 8% (±4% 95% CI) greater using the volume approach than the allomctric approach, and 4% greater (±4%) in the Volgograd region. The greatest difference was for pine dominated stands (-15%) and the least for birch dominated stands (+1%). We also compared the carbon estimated for the 26 Vologda stands utilizing allomctric equations developed for the same genera growing in similar forest types in North America. The North American allometric equations predicted slightly higher carbon content on average as compared to the Russian derived equations (2% ± 4). The data presented suggest that using volumetrically derived carbon estimates provide reasonably accurate estimates of forest carbon.
Article
The lack of growth and yield information for young even-aged western hemlock (Tsuga heterophylla (Raf.) Sarg.)-Sitka spruce (Picea sitchensis (Bong.) Carr.) stands in southeastern Alaska served as the impetus for a long-term stand-density study begun in 1974. The study has followed permanent growth plots in managed stands under various thinning regimes. Between 1974 and 1987, 272 plots were established at 59 locations throughout southeastern Alaska. Remeasurement of the plots occurs every two to four years and will continue until harvest. Additional thinnings will occur on a portion of the plots. Future plans include extending the study through establishment of installations in stand types not currently represented. Once data for an entire rotation are obtained, a comprehensive set of growth and yield tables for various management regimes can be developed. This information will answer questions forest managers have on whether and when to thin a stand, at what level of intensity, and how frequently to enter the stand.
Article
Three scenarios of the Conservation Reserve Program (CRP) were simulated to project carbon (C) pools and fluxes of associated grassland and forestland for the years 1986-2035; and to evaluate the potential to offset greenhouse gas emissions through C. sequestration. The approach was to link land-area enrolments with grassland and forestland C. densities to simulate C. pools and fluxes over 50 years. The CRP began in 1986 and by 1996 consisted of 16.2× 106 ha cropland converted to 14.7× 106 ha grassland and of 1.5× 106 ha forestland. The CRP1 simulated the likely outcome of the CRP as contracts expire in 1996 with the anticipated return of 8.7× 106 ha grassland and of 0.4× 106 ha forestland to crop production. The CRP2 assumed that the CRP continues with no land returning to crop production. The CRP3 was an expansion of the CRP2 to include afforestation of 4× 106 ha new land. Average net annual C gains for the years 1996-2005 were <1, 12, and 16 TgC yr-1 for CRP1, CRP2, and CRP3, respectively. Afforestation of marginal cropland as simulated under CRP3 could provide approximately 15% of the C offset needed to attain the Climate Change Action Plan of reducing greenhouse gas emmissions to their 1990 level by the year 2000 within the United States.
Article
This study provides historical estimates and projections of U.S. carbon sequestered in wood and paper products and compares them to amounts sequestered in U.S. forests. There are large pools of carbon in forests, in wood and paper products in use, and in dumps and landfills. The size of these carbon pools is increasing. Since 1910, an estimated 2.7 Pg (petagrams; × 10 9 metric tons) of carbon have accumulated and currently reside in wood and paper products in use and in dumps and landfills, including net imports. This is notable compared with the current inventory of carbon in forest trees (13.8 Pg) and forest soils (24.7 Pg). On a yearly basis, net sequestration of carbon in U.S. wood and paper products (additions including net imports, minus emissions from decay and burning each year) is projected to increase from 61 Tg/year in 1990 to 74 Tg/year by 2040, while net additions (sequestration) in forests is projected to decrease from 274 to 161 Tg/year. Net sequestration is increasing in products and landfills because of an increase in wood consumption and a decrease in decay in landfills compared with phased-out dumps. If the total projected amount of products required is regarded as fixed, the net carbon sequestration in products and landfills can be increased by 1) shifting product mix to a greater proportion of lignin-containing products, which decay less in landfills; 2) increasing product recycling; 3) increasing product use-life; and 4) increasing landfill CH 4 burning in place of fossil fuels.