Ecological Applications, 1(2), 1991, pp.
182-195
? 1991
by
the
Ecological Society
of
America
NORTHERN
PEATLANDS: ROLE IN THE CARBON CYCLE
AND PROBABLE RESPONSES
TO CLIMATIC WARMING1
EVILLE GoRHAM
Department
of Ecology, Evolution and Behavior,
University of
Minnesota, Minneapolis,
Minnesota
55455 USA
Abstract. Boreal and subarctic
peatlands comprise a carbon pool of 455 Pg that
has
accumulated during the postglacial
period at an average net rate of 0.096 Pg/yr (1 Pg
=
1015 g). Using Clymo's (1984) model,
the current rate is estimated at 0.076 Pg/yr. Long-
term
drainage of these peatlands
is
estimated
to be causing the oxidation to CO2
of
a
little
more than 0.0085 Pg/yr, with combustion
of fuel peat adding st0.026 Pg/yr. Emissions
of
CH4 are estimated to release
n0.046
Pg of carbon annually.
Uncertainties
beset
estimates of both stocks
and fluxes, particularly
with
regard
to
Soviet peatlands. The influence of
water table alterations upon fluxes of both CO2 and CH4
is
in
great need of investigation over
a wide range of peatland environments, especially
in
regions where permafrost melting,
thermokarst erosion, and the development
of thaw lakes
are likely results of climatic warming.
The role of fire in the carbon cycle of peatlands also
deserves increased attention. Finally,
satellite-monitoring of the abundance of open water
in
the peatlands of the West Siberian
Plain and the Hudson/James Bay Lowland is suggested
as
a likely method of detecting early
effects of climatic warming upon boreal and subarctic
peatlands.
Key words: biomass; carbon cycle; carbon
dioxide; climate warming; greenhouse effect; methane;
mires; peatlands.
INTRODUCTION
Peatlands are characteristic of waterlogged
situations
in which, owing to anoxic and cool
conditions a few
centimetres or decimetres beneath
the surface, organic
detritus accumulates, usually on
relatively flat land-
scapes,
to
depths
>
30
or 40
cm
(depending
upon
which
country's
definition
is
accepted)
and often up
to
several
metres. Peat may consist predominantly
of reed, cat-
tail, and sedge remains
where water enriched
in
bases
and nutrients from surrounding
mineral soils perco-
lates through
the surface
peat;
such peatlands are termed
minerotrophic
fens. Where
peatlands
in
wet climates
are domed
above the
surrounding
landscape and inputs
of bases and nutrients to the
peat
surface are derived
solely
from the
atmosphere, the peat
consists primarily
of the remains of
Sphagnum
mosses; these peatlands
are termed
ombrotrophic bogs.
Many flat peatlands
also have an abundance of
Sphagnum
(Kulczynski 1949;
W. D.
Billings, personal communication).
Situations
such as
these
are
prevalent
over
large
areas
in
the boreal
and subarctic
zones
(Gore 1983),
in which
are
locked
up
vast
amounts
of carbon
sequestered
from
the at-
mosphere by photosynthesis
and not
yet
released
by
decomposition.
Human activities, primarily drainage,
have affected
the
carbon
balance of
peatlands
substantially (Armen-
tano and
Menges 1986,
Gorham
1988). Changing
cli-
matic conditions can also be
expected
to affect
greatly
the
balance
between
photosynthesis
and
decomposi-
tion
in
peatlands,
with
"greenhouse"
warming
of
the
climate,
especially at high latitudes
(Post 1990),
the
most likely
cause of
change.
Peatlands
are unusual
in
"greenhouse"
scenarios be-
cause on
the one hand they
sequester
the
major "green-
house" gas,
CO2, from
the
atmosphere
as
peat,
while
on the other hand
they emit
to it
in
large quantities
both
CO2 and the second most
important "green-
house" gas,
CH4 (Moore and
Knowles 1987, 1989).
Research on
these ecosystems
should focus, therefore,
on the
effects
of
climatic warming
upon
two
aspects
of
peatland
ecology and
biogeochemistry: (1)
the balance
between
CO2 fixation
and
release, and
(2)
the balance
between
CH4
production
and
consumption.
The
influence of climatic
warming upon the carbon
cycle
in
peatlands
will
be largely
indirect. Although
rates
of
photosynthesis,
decomposition, and CH4 emis-
sion
may
all be
expected
to increase
directly
with
rising
temperatures
and
longer growing
seasons,
such
effects
are
likely
to be
strongly
overshadowed
by those caused
indirectly
by hydrological
changes, especially alter-
ations
in
the
level ofthe water
table
(Moore
and
Knowles
1989),
induced by climatic
warming. These effects will
probably
include substantial
water table drawdowns
and
peat
oxidation,
in
southerly
regions owing to great-
er
evapotranspiration,
and
in
more
northerly locations,
where
precipitation may increase
appreciably (Grotch
1988), to
melting of the
permafrost. Such melting
will
probably
cause a good deal of
thermokarst erosion
(Billings et
al. 1982) that will lower water
tables
in
I
Manuscript
received 19
March
1990; accepted
1
June
1990; final version
received 6
August 1990.
May
1991
NORTHERN PEATLANDS AND
THE CARBON CYCLE
183
many areas. However,
it will
also cause the formation
of many shallow thaw ponds and lakes. These will
probably release a good deal of CO2 by oxidation of
eroded peat particulates, but the initiation of hydrosere
succession
in
many of them will lead eventually to
renewed peat accumulation (Chapin et al. 1980, Luken
and Billings 1983).
Climatic
warming may renew peat accumulation
in
subarctic peatlands (Gorham 1988, 1990), where it
ceased long ago owing to climatic cooling and the de-
velopment of continuous permafrost (Zoltai and Tar-
nocai 1975, Zoltai and Pollett 1983). It may also shift
peatland formation,
like the
tree line (Miller 1981),
into landscapes even farther north. Therefore, the to-
pography
of those
landscapes
will
govern where
en-
tirely new peat deposits
will
accumulate. The impor-
tance ofthe topographic factor is evident
in
the location
of the two
major peatland areas
in
the
world. The
largest lies on the vast and nearly level West Siberian
Plain between the Ob and Yenisey rivers
in
the USSR
(Neishtadt 1977, Walter 1977, Neustadt 1984). Slopes
may vary
from
0.1-0.8
in
1000
in
wet
"aapa" peatlands
to as much as
4 in
1000
in
less
wet sites.
The
peats are
also
underlain by relatively impermeable
substrates
(Walter 1977). The next largest peatland occupies the
Hudson/James Bay Lowland of Canada, another re-
gion of flat topography
where the
slope
is
commonly
less than
1 in
1000, and where relatively impermeable
marine silt/clays
and other
deposits
favor
waterlogging
(Riley 1982).
In
this review, because more and better information
is available
concerning
carbon
stocks,
rates
of
accu-
mulation,
and
CH4
emissions
in
North American
peat-
lands, attention
will
be focused upon them. Attempts
will
be
made, however,
to
use
available Eurasian data
and
to
produce global
estimates.
CURRENT
CARBON
STOCKS
AND
DISTRIBUTION
There
is far
more carbon
in
the
peat
beneath the
surface than
in
the
vegetation currently growing
on that
surface.
Boreal
and
subarctic
regions
contain the
largest
areas of
peatland, although
some is found
in
more
temperate
and even
tropical parts
of
the
world
(Gore
1983).
Total carbon pool in boreal
and
subarctic
peat
Boreal and subarctic peatlands
are located almost
wholly
in
the
USSR, Canada,
the USA and the
Fen-
noscandian countries,
with a total area
of
346
x
106
ha. Of
that
area
s
II.5
x
106
ha,
or
3.3%,
has been
drained
(Table 1),
and
about 4.4
x
106 ha
(1.3%)
has
been mined for horticultural
peat
and fuel
(Kivinen
and Pakarinen
1981).
The
mean
depth
of boreal and subarctic
peatlands
is
estimated to be 2.3
m
(Table 1). Using
a
mean bulk
density
of
112
g/L
and
a carbon
content
of
51.7%
of
dry
mass
(much higher than the 40%
in
carbohydrates),
both derived
from extensive
Canadian
data sets (Gor-
ham 1988,
1990), we
can then estimate
readily
the total
carbon
in
the dry mass
of boreal
and subarctic
peat,
subtracting
the mined
area, as (3.42
x
1012 m2)
x (2.3
m)
x
(112
x
103
g/m3)
x (0.517)
= 455 x 10'5 g,
or
455 petagrams
(Pg).
This
amounts
to about one-third
of the total
world pool
of soil carbon
(1395 Pg)
esti-
mated by
Post et al. (1982),
and is substantially
greater
than their
estimate of
carbon tied
up
in
moist and
wet
boreal
forest and tundra
(374 Pg,
or 347 Pg
in
Post
et
al.
1985).
The
estimate
of 455
Pg given above
is considerably
higher than
others, for
instance
my own earlier
value
(Gorham
1990) of 180-227
Pg,
and that of 300
Pg
for
global peatlands
by Sjors
(1980). It
is also higher
than
the estimate of
249
Pg
given by
Armentano
and
Menges
(1986)
for northern peatlands,
but
they assumed
an
average
depth
of
only
1
m.
If
a
depth
of
2.3
m
were
used
instead,
their estimate would
rise
to
573
Pg. Oe-
chel (1989)
estimates boreal peatlands
to
contain
only
210 Pg.
A
carbon pool
of 455 Pg
for boreal and
sub-
arctic
peatlands
is
very
substantially greater
than the
global pools
of dead
organic matter
estimated
by Oe-
chel (1989)
for Arctic
tundra, 55 Pg,
and upland
boreal
forest,
88
Pg.
These numbers
would rise to 61 and 159
Pg, respectively,
if
live
biomass were
added.
Carbon
pool
in
vegetation
Total
biomass of above-
plus belowground
vegeta-
tion in
peatlands
is extremely variable,
depending
on
whether
they are forested.
Dry biomass
in
relatively
open peatlands
can be as low as
760 g/m2,
whereas
in
densely
forested peatlands
it can
be
almost
20 times
as
high,
at
13
800
g/m2
(Grigal
et al.
1985).
Assuming
a carbon content of 45% (Olson
et al.
1983),
these
carbon mass values become
342 and 6210
g/m2. The
median
dry biomass
in 14 peatlands
for
which
data
TABLE 1.
Areas
and
depths
ofboreal
and subarctic peatlands.
Area
(106 ha)
Mean depth
Total
Drained (m)
USSR
150*
3.9*
2.511
Canada
I19t
0.1*
2.21
USA
55t
0.6?
2.5#
Fennoscandia
22*
6.9* 1.1**
Total
346
11.5
2.3tt
*
Data
from
Kivinen and
Pakarinen
(1981).
t
Modified from
Tamnocai
(1984);
see
Major
uncertainties,
Carbon
stock,
but with
new
data
for Ontario (Riley 1988).
*
Alaska
(Kivinen
and Pakarinen
1981);
Minnesota,
Mich-
igan,
Wisconsin,
Maine,
New
Hampshire,
Vermont, and
Washington (McKinzie
1982).
?
Calculated
by applying
percent
drainage
in
Minnesota
to
all
of
the states
in
the
previous
footnote
except
Alaska.
11
Estimate
based
on Neustadt
(1984).
1
Data compiled
by E.
Gorham,
J. A.
Janssens,
S.
C. Zoltai,
and
R. S.
Clymo
(unpublished
manuscript).
#
Estimated
by comparison
with Canadian
data.
**
Data from Lappaleinen
(1980).
tt
Weighted
for area.
184
EVILLE
GORHAM
Ecological Applications
Vol.
1,
No.
2
were
compiled by
Bradbury and
Grace (1983)
and Gri-
gal et
al.
(1985)
is 2760
g/m2 (carbon
1240
g/m2),
whereas
the mean is
4430 g/m2
(carbon
1990 g/m2).
The last of
these
carbon numbers
is close to
the esti-
mate by
Olson et al.
(1983) of
2000 g/m2,
whereas
Oechel
(1989)
estimates 2700
g/m2.
Vegetation vs.
peat
On
a
peatland area
basis the
global carbon
pool of
455 Pg
(estimated
above)
amounts to 133
000 g/m2,
as
compared to
z2000
g/m2 tied
up in
vegetation. It
appears,
therefore,
that ;98.5% of
total
peatland car-
bon
occurs
in
the
form of peat,
as against
1.5% in
the
vegetation.
Carbon
fluxes in
boreal and
subarctic
peatlands
Net carbon
flux from
the
atmosphere to
undrained
peatlands
can be
estimated by
adding to the
data pro-
vided
above a figure
for the
annual rate of
increase
in
height.
Unfortunately,
owing to the
slowness with which
peatlands
plants
decompose, current
rates
of
peat ac-
cumulation
cannot be
measured
directly. Some evi-
dence
(Clymo 1984)
suggests that
decay
may
indeed
continue
anaerobically
over
hundreds and
even
thousands
of
years.
Long-term
rates can,
however, be
measured over
thousands
of
years
by
radiocarbon
dat-
ing.
Several
such
estimates for
varying lengths
of
time
in
the
postglacial are
given
in
Table
2;
those of E.
Gor-
ham, J. A.
Janssens,
S. C.
Zoltai, and
R.
S.
Clymo
(unpublished
manuscript)
are derived
wholly
from
138
basal
14C
dates for the
entire
region.
An
overall
height
accumulation
rate
of
0.5
mm/yr
seems both
conser-
vative and reasonable.
Assuming that this
figure ap-
plies
today, the dry-mass carbon
flux
from the
atmo-
sphere
to
undrained and unmined boreal and
subarctic
peatlands
can be calculated as
(3.30
x
1012
m2)
x
(0.0005myr-1)
x
(112
x
103g/m3)
x
(0.517)=0.096
x
1015
g/yr, or
0.096 Pg/yr. On
an area
basis
this
amounts
to 29
gm-2
yr-'.
Divided
into the
total
pool
of
439
Pg
for
these same
peatlands,
a rate
of
0.096
Pg/
yr yields
an
average
age
for
the
peat
deposits
of
4600
yr.
This estimate
of
long-term carbon
storage
at
0.096
Pg/yr
is
very similar to
my earlier
estimate
of
0.091
Pg/yr
(Gorham 1990)
and also to
that of Sjors
(1980),
0.090
Pg/yr,
for
global
peatlands.
Silvola
(1986) sug-
gested
a
higher value,
0.110
Pg/yr,
for
global
peatlands,
and
Armentano and
Menges
(1986)
a lower
range,
0.050-0.075
Pg/yr,
for northern
peatlands.
The
assumption
that the estimated
overall
accretion
rate
of 0.5
mm/yr
applies today
is, however,
unlikely
as a
general
rule.
Clymo
(1984)
has
developed
a model
of
decreasing peat
accumulation over
time,
as
follows:
m
=
p/a)(l-e-at),
where
m
=
accumulated
mass
on an
area
basis
at time
t, p
=
the annual
rate
of
dry
mass
addition to
the anaerobic
peat
mass,
the
catotelm,
after
aerobic
decay
has
occurred
in
the
acrotelm
above,
and
a =
the
fraction of
decomposition
in
the
total
mass of
the
catotelm,
which
continues to
release
CO2
and
CH4
anaerobically and
very
slowly.
For
38
boreal
peat
cores (E.
Gorham, J.
A.
Janssens,
S. C.
Zoltai,
and R. S.
Clymo,
unpublished
manuscript)
whose
diverse
basal
14C
dates
have
been
treated as
coming
from
a
single
peat
profile
(thereby
assuming
constancy
of
p and
a
over both
space
and
time),
p
=
80
g.m-2
yr-' and
a
=
0.00014.
If
these
values
are
applied to an
average peat
depth
of
2.3
m
(Table
1),
and
assumptions
are made of bulk
density
=
112
g/L
and carbon
fraction
=
0.517,
then
the
current carbon
accretion
rate
becomes 23
g.m-2-
yr-1.
Over
a total
area
of
3.30
x
1012
m2
this
becomes
0.076
Pg/yr.
This
same
model,
with the
same
assumptions,
yields
an
average
age
for
these
peat
deposits
of
4300
yr.
Divided into
2.3
m
depth,
this
results
in
an
overall
peat accumu-
lation
rate of
0.53
mm/yr,
very
similar to
that
derived
from
the
data
in
Table 2.
There is
a
further
question,
whether
individual
peat-
lands
are
spreading
laterally
or,
alternatively, are
con-
tracting
in
area.
Although
erosion
and
gullying
do
dam-
age some
peatlands,
most often
those
subjected
to
human
disturbance,
there
is
no
evidence for
major
contractions
in
the area of
undisturbed
peatlands.
The
evidence
for
present
expansion
is
scanty,
and
some
boreal
peatlands
have shown
very
little
increase
in
re-
cent
millennia
(Malmstr6m
1923).
According
to
Sj6rs
(1982)
the
main
period
of
peatland
spreading
(palu-
dification)
was
between
5000
and 2000
yr
BP.
How-
ever, Neustadt
(1984),
taking
the
large
(2268
km2)
Bak-
char
Bog
in
Western
Siberia
as
typical and
assuming
an
accretion rate
of 0.5
mm/yr
(Neyshtadt
et al.
1974),
suggests
that
spreading
has
continued
fairly
rapidly
up
to the
present
time
(Table
3).
During the
early
Holo-
cene
in
west-central
Canada,
fens
and
extensive
peat
TABLE 2.
Long-term
height
accumulation
rates
in
boreal
and
subarctic
peatlands,
calculated as
current
peat
depth
+
basal
age.
Height
accumu-
lation
rate
Location
(mm/yr)
South
Sweden and
North
Germany (median
value
from
Tolonen
1979)
0.70
South
and
central
Finland
(median
value from
Tolonen
1979)
0.75
Northem
Europe
(Aaby
1986)
0.60
Boreal
USSR
(Botch and
Masing
1983)
0.6-0.8
Siberian
palsa
province
(Botch and
Masing
1983)
0.2-0.4
Eurasia
(Zurek
1976)
0.52
Subarctic
Canada
(E. Gorham
et
al.,
unpub-
lished
manuscript)
0.31
Boreal
Canada
(E.
Gorham
et
al.,
unpublished
manuscript)
0.54
Canada
overall
(E.
Gorham
et
al.,
unpublished
manuscript)
0.48
May 1991 NORTHERN PEATLANDS AND THE CARBON CYCLE 185
TABLE 3. Spreading
of the
Bakchar Bog in
Westem
Siberia
during postglacial
time
(Neustadt
1984).
Period (yr
BP) Area (km2)
Increase
(km2)
8000
32
333
6000
365
642
4000
1007
801
2000 1808
460
0
2268
accumulation
occurred only
in the Rocky
Mountain
foothills
and north of
;53330'
N. In middle
and late
Holocene (after
;6000 yr
BP) such fens
expanded
southward,
probably
in
response
to declining
climatic
aridity (Zoltai
and Vitt 1990).
Net carbon
flux to the atmosphere
from drained
bo-
real and subarctic
peatlands
is more difficult
to esti-
mate because
of the extreme
paucity of data.
According
to Silvola (1986),
undisturbed
Finnish peatlands
tend
to release CO2
at the rate of
100-150 mg.m-2
h-1
at
summer
temperatures
of
1
0?-
1
5C. Lowering
the
water
table
in
one such
peatland
from the undrained
range
of
0-10
cm
to the drained range
of 40-60 cm
beneath
the surface increased the
CO2
output
within
a
few weeks
to
300-400
mg.m-2-h-1,
where it remained
for
3
yr.
(Quite
similar results have been reported
in
experi-
mental peat
cores by Moore
and Knowles
1989.)
In
the
year
following drainage,
the organic-matter
equiv-
alent
of
CO2
release was estimated
as
;
700
g*m-2.
yr-1,
or (assuming that
carbon content
=
51.7%)
a carbon
equivalent
of 362
g.m-2 yr-1.
Taking
this
value as characteristic
(drainage
for forestry
accounts
for
t77%
of total
drainage
in
the countries
listed
in
Table 1,
according to Kivinen
and Pakarinen
1981),
the total release
of carbon
by drainage
of
boreal
and
subarctic
peatlands
would
be
(0.115
x
10
12
m2)
x
(362
gm-2-yr-1)
=
0.042
x
1015
g/yr,
or
0.042 Pg/yr.
This
is
a substantial
fraction (55%)
of the
carbon just esti-
mated to be sequestered
currently by
undrained peat-
lands.
However,
releases
over the
longer
term
may
be
very
much
less. According
to Armentano
and
Menges
(1986),
carbon release from
Finnish/Soviet peatlands
drained
for pasturing
and
forestry
for
many
decades
is
only
;30
g-m-2
yr-'.
Drainage
for
crops
is estimated
to
release
217
g
m-2
yr-1.
Applying
these
figures,
weighted
for the areas drained
for
forestry
and
agri-
culture
in Finland and
the
USSR
(the leading
countries
in
total
areas of
peatland
drainage [Kivinen
1980])
yields
a
mean
long-term
carbon
release
rate
of
74
g.m-2 -yr-'.
Taking
this
value,
the
total
release of
car-
bon
by long-term
drainage
ofboreal and subarctic
peat-
lands would be
only
0.0085 Pg/yr, about one-fifth
of
the 0.042
Pg
estimated using
a short-term release
rate.
The
true
release rate should
presumably lie
somewhat
above the lower
estimate.
The
utilization
of peat for
fuel releases
yet
more
carbon
dioxide to the atmosphere.
According
to
Kivi-
nen
(1980),
the
USSR
produces
;100
x 106
tons
(megagrams)
of
fuel peat
(40%
water)
annually,
with
Finland the
only
other significant
boreal-subarctic
pro-
ducer at
1.5
x
106 tons.
Botch
and
Masing
(1983),
however,
claim that
Soviet
fuel
peat production
is only
60-65
x
106 tons/yr
and declining.
Averaging
these
two estimates
for
the USSR,
the
annual
carbon
release
from
fuel
peat is
(82.75
x
1012
g)
x
(0.60
[=
fraction
dry mass])
x
(0.517
[=
fraction carbon])
=
0.026
Pg/
yr.
Net flux
of CH4
from
boreal
and subarctic
peatlands
to the
atmosphere
is
also difficult
to
estimate.
Data
are
scarce,
tend to
be
logarithmically
distributed,
and
do
not
include
fluxes
by ebullition
(Harriss
et al.
1985).
Addition
of tower
and
airplane
measurements,
on a
much
larger
scale than current
flux-chamber
measure-
ments,
may improve
spatial
estimates.
They
are
cur-
rently
being
tested
by
scientists
with NASA
support.
CH4
estimates
are affected
strongly by
temperature
(Crill
et al. 1988,
Whalen and
Reeburgh
1988)
and
depth
of
water
table
(Harriss
et al. 1982,
Sebacher et
al.
1986;
N.
B.
Dise, personal
communication).
Nevertheless,
data compiled
by Crill
et
al. (1988)
(Table
4)
show
rather
little
variation
in
peak
midsummer
carbon
flux-
es
(means
of 96
to
188
mg.m-2
d-1)
over a
range
from
260
to 620
N latitude,
the
lowest value
coming
from
Florida
and
the highest
from
the mountains
of West
Virginia.
Variation
in
water table
depth
is
probably
an
important
confounding
factor
in
these sites.
Distinctly
lower
peak
carbon
flux
values (23
and
46
mg-m-2-d-')
are
given
for northern sites
(650
and 560
N)
by
Whalen
and
Reeburgh (1988)
and
by
Moore and
Knowles
(1987).
By far the
largest
data
base
comes
from Min-
nesota
(Crill
et al.
1988),
and
if we employ
half that
peak
carbon
emission rate
of
155
mgiM-2
.
d-l
for an
average
of 180
d/yr
above
00
(Gorham
1988), CH4-C
emissions
from undrained
boreal and subarctic
peat-
lands would amount
to
(3.30
x
10
12
m2)
x
(0.0775
g-m-2-d-1)
x
(1.80
d)
=
0.046
Pg/yr.
This
is
almost
the
same as
the estimate by
Matthews and
Fung (1987),
using quite
different methods,
for
wetlands between
50?
and
70?
N.
The various carbon
fluxes,
and the
total
carbon
stock,
are
summarized
in
Table 5.
It
appears
that the amount
of carbon released
as
CO2
from northern
peat
deposits
to
the
atmosphere
owing
to
drainage
and
peat
com-
TABLE 4. Carbon
flux as methane
in
midsummer
along
a
latitudinal
gradient
(from
Crill
et al.
1988).
Methane-
Latitude
Number of
carbon
flux
Location
(ON)
samples
(mg
m-2
d-)
Florida
26
11
96
Georgia
30
12 106
W.
Virginia
39
14 188
Minnesota
47
179
155
Alaska
62 13
147
186
EVILLE
GORHAM
Ecological
Applications
Vol.
1,
No.
2
bustion
(0.035 Pg)
is less than
one-half of
the current
amount
added
annually to
the
peat
in
undrained boreal
and
subarctic fens
and bogs (0.076
Pg).
Carbon released
from those
same undrained
peatlands
as CH4,
although
only
61% of the amount
currently
fixed by peat ac-
cumulation,
has a
much greater
climatic
significance.
Multiplied
by a factor
of20 to take
account
ofits greater
effectiveness as a "greenhouse"
gas (Mooney
et al. 1987),
the release
of CH4-C
would be
equivalent
to a release
of
0.92
Pg
of
C02-C,
more than
an order
of magnitude
greater
than the amount
of
C02-C
currently
fixed by
peat deposition,
and
about
26 times that
released by
drainage
and peat-fuel
combustion.
Whether
such
a
multiplication
factor
is justified
has recently
been ques-
tioned
by Lashof and
Ahuja (1990),
who
point out that
the residence
time
of CH4
in
the
atmosphere
(14.4 yr)
is much
shorter than
that of
CO2 (230 yr).
They esti-
mate
that
the
global
warming
potential
of
CH4
may
only
be
-
3.7 times that
of
CO2
on a molar basis.
The
shorter
residence
time of CH4
reflects
its
greater
chem-
ical reactivity
in the atmosphere,
which may have side
effects not presently
calculable.
MAJOR UNCERTAINTIES
IN
ESTIMATES OF
STOcKs
AND FLUXES
The data bases for
both stocks and fluxes are
inad-
equate
in
almost
every way,
partly
because
peatlands
have
received proportionately
little attention
from
ecologists
and biogeochemists
as compared
to forests,
grasslands,
and aquatic
ecosystems.
Fortunately,
peat-
land vegetation
and
landforms
are relatively
similar
in
North America and Eurasia (compare
Sjors
1961,
1963,
Walter
1977, Glaser
et al.
1981, and
Wheeler et al.
1983),
so that generalizations
from
one
continent to
another are
quite
reasonable.
Carbon
stock
Three
major
deficiencies
exist
in
measurements
of
carbon stocks
in
peatlands;
they
concern
estimates
of
both
area and
depth
of
peat,
and also
its
bulk
density.
These deficiencies
can
be illustrated
with reference
to
the two countries
that contain
by
far the
largest
areas
of
peatland,
the
USSR and Canada.
The estimate
of Canadian
peatland
area
(Fig. 1)
is
based
on inventories
taken
by government
agencies
but,
in
northern Canada
especially,
these
inventories
are
often
either
broad-scale
or
lacking (Tarnocai
1984).
In
Canada
peatlands
are defined as having
a
minimum
40-cm
peat depth
(Tarnocai
1984),
whereas
elsewhere
a 30-cm
depth
is
usually
definitive
(Kivinen
and Paka-
rinen
1981).
To
convert,
rather
crudely,
the
Canadian
data
to a 30-cm
depth limit,
I
have added
one-quarter
of the area of Canadian
wetlands
that are shallower
than
40 cm
(Zoltai
1988),
and assumed
their
mean
depth
to
be 35 cm.
The mean
depth
of
Canada's
peatlands
is also not
securely founded,
thousands
of
measurements
being
taken as
representative
of millions
of
hectares
without
any effort
at stratified
random sampling.
My
estimate
of 2.2
m
is lower
than
the
2.7 m calculated from
data
compiled
by Tarnocai
(1984),
in
particular
because
his
estimates
for the
large
areas
of
peatland
in Ontario
and
Quebec
are distinctly
higher than
those given
by J.
L.
Riley (1987a
and personal
communication),
and
Bo-
ville et al.
(n.d.).
The area
of Soviet
peatlands (Fig.
2) is much
more
poorly
known,
or
if
known,
is
not
readily
accessible.
Kivinen
and Pakarinen
(1981)
estimate it,
without
much explanation,
at
150
x
106
ha,
but
according
to
Sabo
(1980)
the total wetland area (excluding
tundra
and forest-tundra)
is 245
x
106
ha (cf.
Neustadt
1984).
The wetland/peatland
quotient resulting
from
these two
estimates
is
1.6,
much
higher
than the
quotient
of 1.1
for Canadian peatlands
(data
of Zoltai 1988).
The
mean depth of
Soviet peatlands
is
also
not well
established.
Neustadt
(1984)
indicates that
in
the
major
peat bogs
it
is "usually
not greater
than
3
to
4
m."
In
the belt of intensive
peat
accumulation
he
states
that
it averages
2.2 m, being
considerably
less (1.0-1.5
m)
in regions
with fewer
peatlands.
I
have, therefore,
es-
timated the overall
mean
depth
conservatively
at
2.5 m.
The bulk
density
of
peat
can
vary widely
both from
place to
place and
within a single peat
core.
I
have used
(Gorham
1988)
a mean
figure
of
112
g/L
for
Canadian
peats,
derived
from extensive
data sets of
Tamocai
(1984),
Boville et al. (n.d.),
Riley (1987a,b),
Riley and
Michaud
(1987),
and
E.
Gorham
and
J.
A.
Janssens,
(unpublished
manuscript).
As
far
as
I
know,
mean data
not been
calculated
for the USSR
or for Fennoscandia,
nor have
I
located any
large data
bases.
During the years
to come it
will be especially.
im-
portant
to
measure,
by
satellite imagery, changes
in
the
area
of
boreal and subarctic
peatlands,
because
climate
change
is
likely
to
destroy
them
in
some regions
while
stimulating their spread
in others.
It
may
be noted as
a final point
that data
on total
live
biomass
(above-
plus belowground)
in boreal
and
TABLE 5. Present
net carbon fluxes to and
from boreal and
subarctic
peatlands,
and the current carbon
stock,
in
peta-
grams
(1 Pg=
1015
g).
Undrained peatlands
(Area
3.30 x 1012
m2)
Carbon flux
(Pg/yr)
Accumulated as organic
carbon
in
peat
Overall
0.096
Current
0.076
Released as CH4 to atmosphere
0.046
Drained peatlands (Area
0.115
x
1012 m2)
Carbon
flux
(Pg/yr)
Released as CO2 by
long-term drainage
0.0085
Released as
CO2
by
fuel combustion
0.026
All unmined peatlands
(Area 3.42
x
1012
m2)
Carbon stock
(Pg)
Deposited as peat
over postglacial time
455
