A reassessment of carbon content in wood: variation within and between 41 North American species

Abstract

At present, 50% (w/w) carbon is widely promulgated as a generic value for wood; however, the literature yields few data and indicates that very little research has actually been done. C contents in heartwood of forty-one softwood and hardwood species were determined. C in kiln-dried hardwood species ranged from 46.27% to 49.97% (w/w), in conifers from 47.21% to 55.2%. The higher C in conifers agrees with their higher lignin content (∼30%, versus ∼20% for hardwoods). Wood-meal samples drilled from discrete early wood and late wood zones of seven of the forty-one species were also investigated. C contents of early woods were invariably higher than those in corresponding late woods, again in agreement with early wood having higher lignin content. Further investigation was made into freshly harvested wood of some species to determine how much volatile C is present, comparing oven-dried wood meal with wood meal dried at ambient temperature over a desiccant. C contents of oven-dried woods were significantly lower, indicating that all past data on C content in oven- or kiln-dried woods may be inaccurate in relation to the true C content of forests. We conclude that C content varies substantially among species as well as within individual trees. Clearly, a 50% generic value is an oversimplification of limited application in relation to global warming and the concept of “carbon credits”.

Full text

Available online at www.sciencedirect.com
Biomass and Bioenergy 25 (2003) 381 – 388
A reassessment of carbon content in wood:
variation within and between 41 North American species
S.H. Lamlom, R.A. Savidge∗
University of New Brunswick, Faculty of Forestry and Environmental Management Fredericton, N.B., E3B 6C2 Canada
Received 4 September 2002; received in revised form 6 January 2003; accepted 23 January 2003
Abstract
At present, 50% (w/w) carbon is widely promulgated as a generic value for wood; however, the literature yields few data
and indicates that very little research has actually been done. C contents in heartwood of forty-one softwood and hardwood
species were determined. C in kiln-dried hardwood species ranged from 46.27% to 49.97% (w/w), in conifers from 47.21% to
55.2%. The higher C in conifers agrees with their higher lignin content (∼30%, versus ∼20% for hardwoods). Wood-meal
samples drilled from discrete early wood and late wood zones of seven of the forty-one species were also investigated.
C contents of early woods were invariably higher than those in corresponding late woods, again in agreement with early wood
having higher lignin content. Further investigation was made into freshly harvested wood of some species to determine how
much volatile C is present, comparing oven-dried wood meal with wood meal dried at ambient temperature over a desiccant.
C contents of oven-dried woods were signicantly lower, indicating that all past data on C content in oven- or kiln-dried
woods may be inaccurate in relation to the true C content of forests. We conclude that C content varies substantially among
species as well as within individual trees. Clearly, a 50% generic value is an oversimplication of limited application in
relation to global warming and the concept of “carbon credits”.
?2003 Elsevier Ltd. All rights reserved.
Keywords: Annual rings; Carbon content; Carbon sequestration; Conifers; Earlywood; Latewood; Elemental analysis; Hardwoods; Wood
1. Introduction
Since the Industrial Revolution, atmospheric CO2
concentration has risen from 280 to 365 ppm, and pre-
dictions are that it could reach 700 ppm by the second
half of the 21st century [1]. Rising atmospheric CO2
contributes to global warming and is expected to alter
the earth’s climate [2–4].
Forests cover more than one third of the world’s
land area and constitute the major terrestrial carbon
∗Corresponding author. Tel.: +1-506-453-4919; fax: +1-506-
453-3538.
E-mail address: savidge@unb.ca (R.A. Savidge).
pool [5,6]. Trees and other forest plants x CO2
through photosynthesis, and all forest organisms
release CO2through respiration [7,8]. Thus, forests
are both sinks and sources for atmospheric CO2. In-
creasing the biomass or carbon content of the world’s
existing forests through improved forest management
and/or decreased harvesting of less than fully devel-
oped trees is a potential option for enhancing seques-
tration of atmospheric CO2and ameliorating global
warming [8,9]. Once wood has formed, its durability
and inertness enable it, and therefore its carbon, to
remain in organic form over very long periods. For
example, coals contain carbon sequestered several
hundred million years ago, and we note here that
0961-9534/03/$ - see front matter ?2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0961-9534(03)00033-3

382 S.H. Lamlom, R.A. Savidge / Biomass and Bioenergy 25 (2003) 381 – 388
carbon can also still be found in imperfectly petri-
ed (“permineralized”) woods more than 200 million
years old. These considerations suggest that accumu-
lation and better management of the long-term dura-
bility of wood products could be another important
means of reducing atmospheric CO2content.
To determine the role of forests in mitigating atmo-
spheric CO2content globally, as a starting point it is
essential to have accurate inventory of carbon content
in forest organic matter. Carbon occurs in innumer-
able forms within forest ecosystems; however, wood
represents the dominant pool wherever trees at normal
stocking density are at sapling stage or larger [8]. At
present, there are actually few research data sets on
carbon content in woods [10–12]. A generic carbon
concentration of 50% (w/w) has been assumed and
widely promulgated [13–19], but other reports sup-
ported by little if any data claimed that carbon con-
tent of wood varies, depending on the species, at least
over a range from 47–59% [17,20–24]. In defense of
the latter reasoning, each kind of wood tends to be
chemically as well as anatomically unique. Therefore,
it would seem reasonable to expect that each could
have a characteristic carbon content [25].
1.1. Variation in carbon content within the
individual tree
Within any particular hardwood or softwood
species, the younger “juvenile” wood produced in the
crown has characteristics that dierentiate it from the
older more “mature” wood produced below the live
crown [26]. Dierent forms of growth stress can also
leave their imprints in tree rings [27,28]. Normal stem
wood is chemically and anatomically dierent from
reaction wood, and depending on the tree species, the
same applies to sap and heartwoods, early and late
woods, and various other kinds of wood [29]. Even
growing the same species in dierent geographical
locations can result in readily detectable dierences in
wood properties [26,30]. The anatomical and chem-
ical compositional dierences arise in spite of the
cambial genotype being constant throughout the tree
[25,31]. Thus, there can be little doubt that the expla-
nation for wood variation resides in the inuence of
the microenvironment on the biochemical processes
that occur in each individual cambial derivative as it
dierentiates into a wood element [31].
Variation in phenotypic traits, whether viewed at the
level of the whole organism, tissue, individual cell or
macromolecular components, is normal to all biologi-
cal systems. This is the classical concept of G ×E=P,
where G is the genotype, E is the environment, and
P is what we see, the phenotype. Carbon content,
whether determined as total C in wood or the propor-
tion of C locked up in any particular macromolecular
class, can be viewed as the phenotype, determined by
an interaction between the genes and the environment.
Notable environmental and climatic changes have
occurred over the last century, and cambium may be
responding to those changes. On the genetic side,
selective harvesting and imposed reforestation with
“improved” trees have altered genetic diversity.
Although there have been more than two centuries of
forestry and wood properties research, evidently not
one study has addressed the question of variation in
total carbon content of wood, as it exists in the forest
in relation to changes in either G or E [8,25]. In order
to know what if any change in carbon has occurred
due to changing E and/or G, it is necessary to have
accurate baseline data. Here we provide methodology
for accurate determination of carbon in wood and
begin to establish that baseline.
2. Materials and methods
Forty-one sawn and planed blocks of kiln-dried
clear heartwood taken from mature boles of softwood
and hardwood North American species (Table 1) were
investigated for their carbon content. These blocks
were obtained from Forintek Canada Corp. (Eastern
Forest Products Laboratory, Ottawa). The dimen-
sion of each block was 6:5mm×102 mm ×63 mm
(transverse ×radial ×tangential sections, respec-
tively). Discrete early and latewood samples were
investigated in only seven of the forty-one, due to
narrowness of growth rings. The seven species in-
vestigated were Abies amabilis,Abies balsamea,
Fraxinus nigra,Fraxinus americana,Larix laric-
ina,Ulmus rubra and Tsuga canadensis. The carbon
content of anatomically characterized permineralised
wood of Araucarioxylon (Triassic, 225 million years
old) obtained from Monument Valley in Utah was
also determined [25]. Woods from freshly harvested
species were also investigated to determine how much

S.H. Lamlom, R.A. Savidge / Biomass and Bioenergy 25 (2003) 381 – 388 383
Table 1
Carbon and hydrogen contents of hardwood and softwood North American species ±SD
Hardwoods Softwoods
Species (kind of wooda;b;c;d;e) C% H% Species (kind of wooda;b;c;d;e)C%H%
Acer macrophyllum Pursh (jc) 49:64 ±0:27 8:52 ±0:21 Abies amabilis (Dougl.) Forbes (jc) 48:55 ±0:99 8:10 ±0:22
Acer negundo L. (mn) 49:34 ±0:53 8:13 ±0:42 Abies balsamea (L.) Mill. (jc) 50:08 ±0:45 7:69 ±0:35
Acer rubrum L. (jc) 48:64 ±0:52 8:38 ±0:36 Chamaecyparis nootkatensis (D. Don) Spach (mn) 52:84 ±0:55 8:30 ±0:15
Acer saccharum Marsh. (cw) 49:32 ±0:19 7:89 ±0:20 Juniperus virginiana L. (cw) 52:14 ±0:88 8:23 ±0:30
Alnus rubra Bong. (cw) 47:70 ±0:12 7:99 ±0:19 Larix laricina (Du Roi) K. Koch (jc) 47:21 ±0:35 7:90 ±0:18
Betula alleghaniensis Britton (cw) 46:27 ±0:33 5:56 ±2:10 Larix occidentalis Nutt. (jc) 47:60 ±0:21 7:90 ±0:20
Betula papyrifera Marsh. (jc) 48:37 ±0:21 7:87 ±0:26 Picea glauca (Moench) Voss (jc) 50:39 ±0:45 7:95 ±0:26
Carya Nutt. (cw) 48:47 ±0:41 8:02 ±0:35 Picea sitchensis (Bong.) Carr. (cw) 49:95 ±0:02 8:24 ±0:09
Fagus grandifolia Ehrh. (jc) 46:60 ±0:39 6:09 ±0:87 Pinus banksiana Lamb. (cw) 50:40 ±0:43 7:63 ±0:33
Fraxinus americana L. (jc) 48:28 ±0:36 7:90 ±0:26 Pinus contorta Dougl. (jc) 50:32 ±0:43 8:05 ±0:46
Fraxinus nigra Marsh. (mw) 47:80 ±0:48 8:02 ±0:13 Pinus ponderosa Laws. (jc) 52:47 ±0:38 8:34 ±0:34
Juglans cinerea L. (cw) 48:53 ±0:36 7:69 ±0:68 Pinus resinosa Ait. (cw) 53:28 ±0:33 8:74 ±0:07
Juglans nigra L. (cw) 49:17 ±0:12 7:70 ±0:03 Pinus strobus L. (jc) 49:74 ±0:16 8:25 ±0:25
Platanus occidentalis L. (jc) 49:97 ±0:82 8:32 ±0:16 Pseudotsuga menziesii (Mirb.) Franco (cw) 50:50 ±0:36 8:25 ±0:10
Populus tremuloides Michx. (jc) 47:09 ±0:75 6:28 ±1:14 Thuja occidentalis L. (jc) 51:72 ±0:17 8:09 ±0:18
Populus trichocarpa Torr. & Gray (jc) 49:25 ±0:25 8:29 ±0:18 Thuja plicata Donn (mn) 51:54 ±0:38 8:16 ±0:27
Prunus serotina Ehrh. (jc) 49:53 ±0:18 8:00 ±0:34 Tsuga canadensis (L.) Carr. (jc) 50:33 ±0:32 7:63 ±0:47
Quercus alba L. (mw) 49:57 ±0:22 7:64 ±0:25 Tsuga heterophylla (Raf.) Sarg. (mw) 50:60 ±0:45 7:85 ±0:33
Quercus rubra L. (jc) 49:63 ±0:32 8:14 ±0:29 Sequoiadendron giganteum (Lindl.) Bucholz (hwc, mn) 55:16 ±0:52 8:12 ±0:07
Salix L. (cw) 49:05 ±0:58 8:26 ±0:28 Sequoiadendron giganteum (swd;mn) 54:66 ±0:27 8:50 ±0:08
Tilia americana L. (cw) 46:43 ±0:17 6:48 ±0:61 Sequoiadendron giganteum (tze;mn) 52:52 ±0:27 7:77 ±0:09
Ulmus L. (jc) 46:32 ±0:27 5:67 ±0:26
aExcepting where noted for S. giganteum, all specimens were clear heartwood.
bHeartwood was sub-divided as follows based on examination of transverse surfaces: jc, juvenile core wood containing annual rings ¿5 mm in radial width and
having pronounced curvature (i.e., obviously cut from near the pith); cw, core wood containing annual rings having some curvature and between 2–4 mm radial
width; mn, mature outer wood having non-curved narrow (¡2 mm radial width) annual rings; mw, mature outer wood having non-curved wide (¿2 mm radial
width) annual rings.
cHeartwood.
dSapwood.
eTransitional between heartwood and sapwood.

384 S.H. Lamlom, R.A. Savidge / Biomass and Bioenergy 25 (2003) 381 – 388
volatile carbon exists, comparing oven-dried wood
powder with that dried at ambient temperature.
2.1. Preparation of wood for analysis
Each wood block was shaved to a depth of 1–2 mm
immediately before sampling, using a clean sharp ra-
zor blade, in order to expose a non-oxidized transverse
surface. An electric drill with small diameter (0:2mm)
bit was used to obtain wood powder, drilling into the
wood to a depth of 3–4 mm. The powder was trans-
ferred to a glass vial, capped with aluminum foil, and
dried in a vacuum desiccator over indicator silica gel.
In the case of Araucarioxylon, small pieces of solid
material previously established to be anatomically re-
solvable [25] were pulverized using mortar and pestle.
To determine the number of days needed to reach
equilibrium dryness, four species (Acer negundo,
Abies amabilis,Juniperus virginiana, and Picea
sitchensis were randomly selected. Each wood pow-
der (∼10 mg) in a glass vial was covered with foil,
weighed using an analytical balance, and then placed
in a vacuum desiccator over desiccant and subjected
to reduced pressure for 10 min, when the stopcock
was closed. The samples were left in the desiccator
until reweighing the following day, then returned to
the desiccator. Only 10 min of vacuum application
was used each day to minimize loss of volatile or-
ganics. The weights of all samples were observed to
stabilize before ten days; thus, in subsequent research
all samples were dried as described over 10 days be-
fore analysis. Oven-dried samples were prepared in
the same way as those dried at ambient temperature
with the exception that the wood powder was left in
a93
◦C oven for one week.
2.2. Analytical method
Each sample (∼1 mg) of dry wood powder for
elemental analysis was weighed into a clean, dry tin
container (∼33 mg tin, 20 mm dia. circle crimped
into a 5 ×9 mm cup, CE Elantech, Inc.) using a Cahn
C-30 microbalance (calibrated precision 0:001 mg).
The tin containers were pre-washed with double all
glass-distilled water followed by two washes with
analytical grade acetone, then vacuum desiccated
overnight to ensure that they were completely dry.
A Carlo Erba CHN 1500 elemental analyser was
used to quantify total carbon, hydrogen and nitrogen.
Full gas chromatographic resolution of CO2,N
2and
H2O was achieved under the following conditions:
column length, 3 m; column diameter, 6 ×4mm
(OD/ID); packing material, Porapak QS, 50–80 mesh;
UHP helium ow rate, 85 ml=min; helium reference
ow rate, 40 ml=min; gas chromatograph oven tem-
perature, 90◦C; lament temperature (thermoconduc-
tivity detector), 190◦C. In the oxidation furnace, the
combustion products passed through a 12 cm layer of
chromium trioxide (Cr2O3) followed by a 6 cm layer
of silver coated cobalt oxide separated by a few mm
of quartz (silica) wool, all packed within a vertical
clear quartz tube (45 cm long, 14 mm i.d., 18 mm
o.d., ThermoQuest). In the reduction furnace the
mixture of combustion products (CO2,N
2,NO
xand
water) passed through a second quartz tube fully
packed with metallic copper to scrub oxygen and
reduce any nitrous oxides to nitrogen (N2).
2.3. Calibration of the instrument
Under the described conditions, a calibration curve
generated using high purity crystalline L-leucine
(Sigma-Aldrich) yielded the linear regression
YC=6:492 ×106X+ 65809(R2=0:996), where
YCis carbon peak area and Xis unknown carbon
content in milligrams (Fig. 1). The hydrogen per-
centage was calculated from the linear regression
YH= 149:07X−1:45(R2=0:989), where YHis hydro-
gen peak area and Xis unknown hydrogen content in
milligrams (Fig. 2). The linear regression for nitrogen
was YN=2:380 ×106X−1578:937 (R2=0:996),
where YNis nitrogen peak area and Xis unknown ni-
trogen content in milligrams. Glucose and cellobiose
standards (Sigma-Aldrich Canada Ltd) were also in-
vestigated, and their C and H contents were found
to be in good agreement with the leucine calibration
curves.
2.4. Statistical analyses
The mean of at least three replicates per sample
and the standard deviation for those replicates were
calculated. Where the standard deviation for carbon
was greater than 0.6% (w/w), more replicates were
analysed. We tested variation in carbon contents with

S.H. Lamlom, R.A. Savidge / Biomass and Bioenergy 25 (2003) 381 – 388 385
Carbon Content (mg)
Integrated Peak Area Units ( × 106)
0
4
8
12
16
20
0 0.6 1.2 1.8 2.4 3
Y = (6.492 × 106)X + 65809
Fig. 1. L-Leucine calibration curve used to calculate carbon content.
Hydrogen Content (mg)
Integrated Peak Area Units ( × 10
5
)
0
15
30
45
60
75
0 0.1 0.2 0.3 0. 4 0.5
Y = 149.07X – 1.45
Fig. 2. L-Leucine calibration curve used to calculate hydrogen content.
a two-way analysis of variance (=0:001), the factors
being species of tree and method of drying (ambient
versus oven-dried).
3. Results
Analyses of heartwood of 22 hardwood species
showed that the carbon content ranged from 46.27%
to 49.97% (w/w). In contrast, it ranged from 47.21 to
55.2% in heartwood of 19 softwood species (Table 1).
Carbon content of early wood was higher than that
in corresponding late wood in all species (Fig. 3).
The data of Fig. 4indicate that carbon contents of
oven-dried woods were invariably lower than those of
ambient-temperature desiccated woods in the studied
species (two-way ANOVA: species eect: F=73:08,
P=0:000, df = 7; drying treatment: F= 131:92,
P=0:000, df = 1; interaction: F=10:30, P=0:000,
df = 7; error MS = 0:226, df = 49). Hydrogen con-
tents ranged from 5.56% to 8.32% in hardwoods and
from 7.63% to 8.74% in softwoods. Nitrogen was also
analysed, but its content never exceeded trace levels.
In most species the wood powder produced by
drilling yielded only small variation between replicate
analyses (n= 3, minimum). However, for Populus
tremuloides,Platanus occidentalis,Abies amabilis
and Juniperus virginiana standard deviations as great
as 1% were obtained. The explanation for the greater
variability between replicate samples of wood pow-
der of some species is still under investigation. In
some cases grinding wood particles into liquid nitro-
gen solved this problem. For example, in Ulmus the
carbon content was 48:11 ±2:25 for coarsely pow-
dered samples. In contrast, it was 46:32 ±0:27 for
samples nely powdered in liquid nitrogen by means
of mortar and pestle.
The carbon and hydrogen contents of Araucar-
ioxylon permineralized wood were 1:61 ±1:07%
and 0:28 ±0:10%, respectively, in agreement with

386 S.H. Lamlom, R.A. Savidge / Biomass and Bioenergy 25 (2003) 381 – 388
F. nigra
F.americana
Ulmus
A
.amabilis
A.balsamea
T. canadensis
L.laricina
C (% w/w)
43
44
45
46
47
48
49
50
51
Early wood
Late wood
Fig. 3. Carbon contents of late wood and early wood in the seven
studied species. The error bars represent standard deviations.
A.rubrum
A.saccharum
B
.alleghaniensis
F.grandifolia
F.a
mericana
P.strobus
T.occidentalis
(sapwood)
T.occidentalis
(heartwood)
C (% w/w)
42
44
46
48
50
52
54
Desiccated wood
Oven-dried wood
Fig. 4. Carbon contents of oven-dried (93◦C, 1 week) versus
desiccated woods. The error bars represent standard deviations.
cellulose microbrils still being readily detected in the
material [25].
4. Discussion
Thorough investigation of the literature dealing with
carbon in wood revealed that there are actually limited
research data, and what exist are not consistent. For
instance, in a chapter of a book entitled “The chem-
istry of cellulose and wood”, Nikitin [33] reported
that completely dry wood, dried at 105◦C, had a very
similar elementary composition for all species, con-
taining about 49.5% C. Wenzl [13] also stated that all
species had similar elemental composition; absolutely
dry wood of any species containing about 50% car-
bon. Prakash and Murray [34] and Corder [35] stated
that although wood composition may vary from one
species to another and even within a species, depend-
ing on the section of the tree, its age, and ecologi-
cal conditions, dierent species of wood nevertheless
show remarkable uniformity in their elemental com-
position. Thus, recent research dealing with carbon in
trees has assumed the 50% value to be correct, al-
though other values have also been used. For exam-
ple, Clifton et al. [36] assigned 51.4% carbon to wood
whereas Elliott [37] stated that wood has 52% car-
bon. However, those authors gave no source for their
values.
All past data on carbon content appear to have
been based on oven- or kiln-dried woods, but our data
on oven-dried versus desiccated woods indicate that
earlier estimates have biased downward the true car-
bon content that is locked up in forests. The two-way
ANOVA output conrmed that carbon contents sig-
nicantly varied among species as well as by method
of sample preparation; however, our data also show
that the inuence of drying (oven vs. ambient) on the
mean carbon contents depends on what species is un-
der investigation. Earlier investigations concentrated
on wood primarily as an energy resource, and it was
not uncommon for wood to be mixed with bark and
dried at high temperature, ignoring the volatile mat-
ters present in all woods. During our research, larger
wood particles were ground into a ne powder using
liquid nitrogen, and emphasis was placed on achieving
sucient homogeneity in the powder to produce small
standard deviation among replicates. Carbon and hy-
drogen percentages were analysed for both coarse
(¿1 mm) and ne (60:3 mm) powdered samples,
and higher standard deviations attended the coarse
preparations. Thus, our experience indicates that

S.H. Lamlom, R.A. Savidge / Biomass and Bioenergy 25 (2003) 381 – 388 387
accurate estimates can only be achieved by reducing
wood to particle sizes 60:3 mm and, ideally, much
smaller. Others evidently have made no eort to grind
wood samples to ne consistency [10], and carbon
contents having standard deviations as high as 3.2%
have been reported [11].
Our results, based exclusively on bole wood, clearly
indicate that carbon content varies among species as
well as within individual trees (for example early
vs. late woods). The carbon content of softwoods
species is generally higher than that of hardwoods,
in agreement with softwood lignin content being
approximately 10% higher than that of hardwoods
[25] of all the macromolecules making up wood,
lignin has the highest percentage carbon [25,33,38].
The carbon content of early wood is higher than that
of late wood. Late wood is quite consistently higher
in cellulose and lower in lignin than early wood
[25,30,38]. Variation between early and late wood
is expected due to the fact that wood exhibits such
chemical variability within a growth layer [30]. For
example, Andrews [32] found greater dierences in
chemical composition between early and late wood
within an annual ring, than between sapwood and
heartwood of the same Douglas-r tree.
Our ndings revealed a range from 46.27% to
55.2% in the carbon content of North American trees,
wood in mature stems of hardwood species ranging
from 46.27% to 49.97% (average 48.41%), and that
in softwood species from 47.21% to 55.20% (average
51.05%). It is probable that the higher carbon con-
tent in softwoods will be found to apply generally,
because softwoods in general have approximately
10% more lignin than hardwoods [25].
A 1% dierence in carbon content conceivably
could have a signicant impact on wood and pulp
industries in relation to allocation of carbon credits
within the Kyoto Protocol. The uncertainty (or, pre-
cision error) associated with our method was 0.5%
(0.1% weighing, 0.4% leucine standard curve), and
our observed dierences in carbon of 9% therefore
appear quite important. It could be argued that,
all other factors being equal, additional carbon
storage capacity per unit mass exists in softwood
forests. However, this would be simplistic because
many hardwood species have wood densities above
0:6gcm
−3whereas softwoods in general are well be-
low 0:6gcm
−3[25,30]. Thus, high-density hardwood
species, although having lower carbon content per
unit mass than softwoods, will nevertheless contain
the greater quantity of carbon per unit volume. Even
disadvantaged hardwood species such as poplar, that
have less than 50% carbon and also have low-density
wood, if suciently fast growing conceivably could
sequester more carbon than softwoods within a grow-
ing season. To estimate carbon content of forest
stands, it is necessary to take into consideration not
only the several kinds of wood within trees [39]
but also stocking density (e.g. number of trees per
hectare by age volume class). It is apparent from such
considerations that accurate carbon inventories and
management of forest ecosystem carbon pools will
require much greater attention to detail than tradition-
ally has been addressed by foresters. It is clear that
much more research is needed, but from the preced-
ing clarication, there is no doubt that a 50% generic
value for carbon content is an oversimplication of
limited application in relation to global warming and
understanding the role of the forest as a carbon sink.
References
[1] Houghton JT, Meiro-Filho LG, Callander BA, Harris N,
Kattenburg A, Maskell K. Climate change 1995—the science
of climate change. Second Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge, UK:
Cambridge University Press, 1995.
[2] Bazzaz FA. The response of natural ecosystems to the
rising global CO2levels. Annual Review of Ecology and
Systematics 1990;21:167–96.
[3] Sedjo RA. The global carbon cycle. Are forests the missing
sink? Journal of Forestry 1990;88(10):33–4.
[4] Bazzaz FA, Fajer ED. Plant life in a CO2-rich world.
Scientic American 1992;266:68–74.
[5] Melillo JM, Challaghan TV, Woodward FI, Salati E, Sinha
SK. Eects on ecosystems. In: Houghton JT, Jenkins GJ,
Ephraums JJ, editors. Climate change. The IPCC Scientic
Assessment. Cambridge: Cambridge University press, 1990.
p. 283–310.
[6] Roberntz P, Sune L. Eects of long-term CO2enrichment and
nutrient availability in Norway spruce. II. Foliar chemistry.
Trees 1999;14:17–27.
[7] Musselman RC, Fox DG. A review of the role of temperate
forests in the global CO2balance. Journal of the Air and
Waste Management Association 1991;41(6):798–807.
[8] Savidge RA. Forest science & technology to reduce
atmospheric greenhouse gases—an overview, with emphasis
on carbon in Canada’s forests. In: Tsang KT, editor.
Proceedings climate change 2: Canadian Technology

388 S.H. Lamlom, R.A. Savidge / Biomass and Bioenergy 25 (2003) 381 – 388
Development Conference, Section 3A. Toronto, Canada:
Canadian Nuclear Society, 2001. p. 3–22.
[9] Sedjo RA, Solomon AM. Climate and forests. In: Norman
Rosenburg, et al., editors. Warming: abatement and
adaptation. Resource for the Future. Washington, DC, 1989.
p. 105–19.
[10] Mingle JG, Boubel RW. Proximate fuel analysis of some
western wood and bark. Wood Science 1968;1(1):29–36.
[11] Reichle DE, Dinger BE, Edwards NT, Harris WF, Sollins P.
Carbon ow and storage in a forest ecosystem. In: Woodwell
GM, Pecan EV, editors, AEC Symposium, Carbon and the
biosphere, Series No. 30, TN. Conf - 720510. Technical
Information Center, Oak ridge, Tennessee, 1973. p. 345–65.
[12] Chow P, Rolfe GL. Carbon and hydrogen contents of
short-rotation biomass of ve hardwood species. Wood and
Fiber Science 1989;21(1):30–6.
[13] Wenzl HF. The chemical technology of wood. New York:
Academic Press, 1970.
[14] Karchesy J, Koch P. Energy production from hardwoods
growing on southern pine sites. US Department of agriculture.
Forest service. General Technical Report SO-24, 1979.
[15] Sedjo RA. Forests. A tool to moderate global warming?
Environment 1989;31:14–20.
[16] Dewar RC, Cannell MGR. Carbon sequestration in the trees,
products and soils of forest plantations: an analysis using UK
examples. Tree Physiology 1992;11:49–71.
[17] Hollinger DY, Maclaren JP, Beets PN, Turland J. Carbon
sequestration by New Zealand’s Plantation forests. New
Zealand Journal of Forestry Science 1993;23(2):194–208.
[18] Matthews G. The carbon content of trees. Forestry
Commission Technical Paper 4. Forestry commission,
Edinburgh, Scotland, 1993.
[19] Thuille A, Buchmann N, Schulz ED. Tree Physiology
2000;20:849–57.
[20] Fryling GR. Combustion engineering. Combustion
engineering, INC, 1966.
[21] Arola RA. Wood fuels—How do they stack up?. In: Energy
and the wood products industry. Forest Products Research
Society, Madison, Wisconsin, 1976. p. 34–41.
[22] Pingrey DW. Forest products energy overview. In: Energy
and the wood products industry. Forest Products Research
Society, Madison, Wisconsin, 1979. p. 1–16.
[23] Tillman DA, Rossi AJ, Kitto WD. Wood combustion:
principles, processes, and economics. Orlando: Academic
press, 1981. p. 43.
[24] Ragland KW, Aerts DJ, Baker AJ. Properties of wood
for combustion analysis. Bioresource Technology 1991;37:
161–8.
[25] Savidge RA. Biochemistry of seasonal cambial growth and
wood formation-an overview of the challenges. In: Savidge
RA, Barnett JR, Napier R, editors. Cell and molecular biology
of wood formation. Bios Scientic Publishers Limited,
2000.
[26] Zobel BJ, Sprague JR. Juvenile wood in forest trees. Berlin
Heidelberg: Springer, 1998.
[27] Douglass AE. Evidence of climatic eects in the annual rings
of trees. Ecology 1920;1:24–32.
[28] Bradley RS. Paleoclimatology: reconstructing climates of the
quaternary. 2nd ed. Harcourt/Academic, 1999.
[29] Timell TE. Compression wood in gymnosperms. Heidelberg:
Springer, 1986.
[30] Zobel BJ, van Buijtenen JP. Wood variation. Berlin
Heidelberg: Springer, 1989.
[31] Savidge RA. Xylogenesis, genetic and environmental
regulation—a review. IAWA Journal 1996;17:269–310.
[32] Andrews EK. Impact of ber morphology and chemical
composition on the Kraft process and subsequent handsheet
properties. R&D Conference TAPPI Proceedings, Raleigh,
North Carolina, 1986. p. 111–9.
[33] Nikitin NI. The chemistry of cellulose and wood. Israel
Program for Scientic Translations, Jerusalem, 1966.
[34] Prakash CB, Murray FE. A review on wood waste burning.
Pulp and Paper Magazine of Canada 1972;73(7):70–5.
[35] Corder SE. Wood and bark as fuel. Research Bulletin
14. Forest Research laboratory Oregon State University,
Corvallis, Oregon, 1973.
[36] Clifton DS, Bulpitt WS, Bircheld JL, Brown ML, Chiang TI,
Day SW, Dixit BS, Gibson NS, McGowan TF. A feasibility
study for wood energy utilization in Georgia. Georgia
Institute of Technology, Final Report, Project A-2140, August
1979.
[37] Elliott RN. Wood combustion. In: Levi MP, O’Grady MJ,
editors. Decisionmaker’s guide to wood fuel for small
industrial energy users. SERI/TR-8234-1. Solar Energy
Research Institute, Golden, Colorado, February 1980. p. 31.
[38] Browning BL. The chemistry of wood. London, New York:
Interscience, 1963.
[39] Barton GM. Denition of biomass samples involving wood,
bark and foliage. Biomass 1984;4:311–4.