Comparison of carbon and water vapor exchange of forest and grassland in permafrost regions, Central Yakutia, Russia


LopezC.,M.L.; Gerasimov,E.; Machimura,T.; Takakai,F.; Iwahana,G.; Fedorov,A.N.; Fukuda,M.

Agricultural and Forest Meteorology 148(12): 1968-1977

2008


Boreal grasslands have been largely neglected in carbon and water vapor flux models despite being originated by past global climate changes. Therefore in this study, meteorological conditions, water vapor and CO<sub>2</sub> fluxes were measured by the eddy correlation technique simultaneously in a larch forest and alas ecosystem (grassland thermokarst depression) in Central Yakutia, eastern Siberia, during the growing season of 2006 (approximately 100 days, May 23rd-August 31st). The alas ecosystem was a carbon sink (-1.38 tC ha<sup>-1</sup>) but had a 60% lower carbon sequestration capacity than the surrounding larch forest (-3.44 tC ha<sup>-1</sup>) during the study period. Despite this large difference in carbon exchange, water loss from the alas ecosystem (118 mm) was only 13% lower than that from the forest ecosystem (136 mm). Water vapor flux measured in the alas was higher under similar environmental conditions when the source was the lake water than when the source was the grassland. This supports the theory that lake evaporation contributes significantly to the evaporation from the alas as indicated also by the lake water level constant decrease during the growing season. Mid-summer forest and alas mean evapotranspiration was 1.4 and 1.2 mm d<sup>-1</sup> respectively. Mean daily canopy conductance was higher in the forest than in the alas (3.8 and 2.4 mm s<sup>-1</sup>, respectively) as expected due to differences in canopy architecture at each site. In this study a rough estimate of the NEE of grassland in Central Yakutia shows an underestimation of 0.9x10<sup>-3</sup> Pg if this area is considered as forested, as most regional models do. Our results suggest that a more detail analysis of distinctive areas within the territory of eastern Siberia is needed in order to obtain a better understanding of carbon and water fluxes from this immense boreal region. Furthermore, if the present global warming evokes landscape change from forest to grassland, the carbon sink capacity of this boreal region could be significantly reduced.

AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
Agricultural
and
Forest
Meteorology
available
at
www.sciencedirect.com
Cr
EI
C1‘71r'173
1
•••
ScienceDirect
journal
homepage:
www.elsevier.com/locate/agrformet
Comparison
of
carbon
and
water
vapor
exchange
of
forest
and
grassland
in
permafrost
regions,
Central
Yakutia,
Russia
M.L.
Lopez
C.`",
E.
Gerasimou
b
,
T.
Machimura
C
,
F.
Takakai
d
,
G.
Iwahana
e
,
A.N.
Fedorou
b
,
M.
Fukudaf
a
United
Graduate
School
of
Agricultural
Sciences,
Iwate
University
Morioka,
Japan
b
Permafrost
Institute,
Siberian
Branch,
Ras,
Yakutsk,
Russia
Graduate
School
of
Engineering,
Oosaka
University
Oosaka,
Japan
d
Graduate
School
of
Agriculture,
Hokkaido
University
Sapporo,
Japan
e
Graduate
School
of
Engineering,
Hokkaido
University
Sapporo,
Japan
f
Institute
of
Low
Temperature,
Hokkaido
University
Sapporo,
Japan
ARTICLE
INFO
ABSTRACT
Article
history:
Received
31
October
2007
Received
in
revised
form
8
September
2008
Accepted
24
September
2008
Keywords:
Carbon
balance
Evapotranspiration
Global
warming
Grassland
Larch
forest
Boreal
grasslands
have
been
largely
neglected
in
carbon
and
water
vapor
flux
models
despite
being
originated
by
past
global
climate
changes.
Therefore
in
this
study,
meteorological
conditions,
water
vapor
and
CO
2
fluxes
were
measured
by
the
eddy
correlation
technique
simultaneously
in
a
larch
forest
and
alas
ecosystem
(grassland
thermokarst
depression)
in
Central
Yakutia,
eastern
Siberia,
during
the
growing
season
of
2006
(approximately
100
days,
May
23rd-August
31st).
The
alas
ecosystem
was
a
carbon
sink
(-1.38
tC
ha
-1
)
but
had
a
60%
lower
carbon
sequestration
capacity
than
the
surrounding
larch
forest
(-3.44
tC
ha
1
)
during
the
study
period.
Despite
this
large
difference
in
carbon
exchange,
water
loss
from
the
alas
ecosystem
(118
mm)
was
only
13%
lower
than
that
from
the
forest
ecosystem
(136
mm).
Water
vapor
flux
measured
in
the
alas
was
higher
under
similar
environmental
conditions
when
the
source
was
the
lake
water
than
when
the
source
was
the
grassland.
This
supports
the
theory
that
lake
evaporation
contributes
significantly
to
the
evaporation
from
the
alas
as
indicated
also
by
the
lake
water
level
constant
decrease
during
the
growing
season.
Mid-summer
forest
and
alas
mean
evapotranspiration
was
1.4
and
1.2
mm
d
-1
respectively.
Mean
daily
canopy
conductance
was
higher
in
the
forest
than
in
the
alas
(3.8
and
2.4
mm
s
-1
,
respectively)
as
expected
due
to
differences
in
canopy
architecture
at
each
site.
In
this
study
a
rough
estimate
of
the
NEE
of
grassland
in
Central
Yakutia
shows
an
underestimation
of
0.9
x
10
-3
Pg
if
this
area
is
considered
as
forested,
as
most
regional
models
do.
Our
results
suggest
that
a
more
detail
analysis
of
distinctive
areas
within
the
territory
of
eastern
Siberia
is
needed
in
order
to
obtain
a
better
understanding
of
carbon
and
water
fluxes
from
this
immense
boreal
region.
Furthermore,
if
the
present
global
warming
evokes
landscape
change
from
forest
to
grassland,
the
carbon
sink
capacity
of
this
boreal
region
could
be
significantly
reduced.
©
2008
Elsevier
B.V.
All
rights
reserved.
*
Corresponding
author
at:
United
Graduate
School
of
Agricultural
Sciences,
Morioka-shi,
Ueda
3
Choume
18-8,
020-8550,
Iwate
University,
Morioka,
Iwate
Prefecture,
Japan.
Tel.:
+81
19
621
6884;
fax:
+81
19
621
6248.
E-mail
address:
larry@iwate-u.ac.jp
(M.L.
Lopez
C.).
0168-1923/$
-
see
front
matter
©
2008
Elsevier
B.V.
All
rights
reserved.
doi:10.1016/j.agrformet.2008.09.013
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
1969
1.
Introduction
The
eastern
Siberian
Taiga
is
an
immense
forested
area
comprising
several
tree
species
and
extensively
dominated
by
larch
(Larix
cajanderi)
stands
(Shvidenko
and
Nilsson,
1994).
This
area
is
underlain
by
continuous
permafrost,
where
methane
(Brouchkov
et
al.,
2004)
and
salt
(Lopez
et
al.,
2007a)
concentrations
are
high
and
is
considered
an
important
global
carbon
reservoir
(Dutta
et
al.,
2006).
Located
in
eastern
Siberia
is
the
area
of
lowland
Central
Yakutia,
an
area
of
approximately
2.4
million
ha,
where
the
underlying
perma-
frost
is
characterized
by
a
high
concentration
of
ice
wedges,
6-
40
m
long
and
more
than
90%
ice
content.
An
important
characteristic
of
this
region
is
the
presence,
intermingled
within
the
forest,
of
thermokast
depressions
(grassland
known
as
alas).
These
thermokarst
depressions
developed
after
permafrost
degradation
that
occurred
during
the
climatic
warming
of
the
early
Holocene
(Katamura
et
al.,
2006)
and
middle
Holocene
(Kachurin,
1962).
An
alas
ecosystem
is
the
last
stage
of
the
thermokarst
phenomenon
(Czudek
and
Demek,
1970).
The
trigger
of
this
thermokarst
phenomenon
is
the
thawing
of
the
upper
permafrost
and
especially
of
ice-rich
permafrost,
which
causes
ground
subsidence.
The
next
stage
is
water
pool
formation,
tree
fall
and
more
importantly,
release
of
salts
from
the
ion-rich
permafrost
(Lopez
et
al.,
2007a),
which,
due
to
the
predominance
of
evaporation
over
precipitation,
brings
salt
up
to
the
surface
making
reforesta-
tion
impossible
(Desyatkin,
1993).
The
total
number
of
alas
is
16,000
and
the
area
covered
is
440
thousand
ha,
which
represents
18%
of
lowland
Central
Yakutia
(Bosikov,
1991).
They
range
in
size
from
about
0.1
to
15
km
in
diameter
and
a
3
to
40
m
in
depth
relative
to
the
surrounding
terrain.
The
area
consists
of
a
group
of
Lena
River
terraces
with
elevations
of
200-220
m
a.s.l.
Ice
wedges
are
spread
over
20%
of
the
territory
of
Central
Yakutia
with
thickness
of
up
to
20-25
m
(Fedorov
et
al.,
1991).
Historic
data
from
a
meteorological
station
in
Yakutsk
show
that
higher
precipitation
and
temperature
values
have
been
registered
in
winter
rather
than
summer
for
the
last
100
years.
This
creates
conditions
for
warmer
ground
tempera-
tures
in
winter
because
of
longer
and
thicker
snow
cover
(Smith,
1975)
that
can
accelerate
permafrost
degradation
(Payette
and
Delwaide,
2000).
There
is
evidence
that
the
present
global
warming
is
accelerating
the
instability
of
other
permafrost
regions
(Osterkamp
and
Romanovsky,
1999;
Jorgenson
et
al.,
2006),
increasing
the
possibility
that
thermo-
karst
grassland
depressions
area
could
expand.
In
the last
decade,
studies
of
fluxes
in
eastern
Siberia
have
improved
our
understanding
of
the
processes
and
feedback
that
might
take
place
as
a
consequence
of
climate
change.
Based
on
measurements
of
forest
water
flux
in
Central
Yakutia,
maximum
evaporation
from
larch
forest
can
reach
values
between
2.5
and
3.0
mm
d'
(Baldocchi
et
al.,
2000;
Ohta
et
al.,
2001),
evenly
distributed
between
tree
canopy
and
understorey
transpiration
(Kelliher
et
al.,
1997;
Lopez
et
al.,
2007b).
Annual
forest
evaporation
closely
approaches
annual
precipitation
(230
mm
y").
Meanwhile,
carbon
flux
measure-
ments
(Hollinger
et
al.,
1998;
Dolman
et
al.,
2004;
Machimura
et
al.,
2005)
have
revealed
a
net
carbon
sequestration
that
ranges
from
0.9
to
2.3
tC
ha
-1
,
which
is
a
reflection
of
the
strong
inter-annual
climatic
variability.
The
Taiga
is
spatially
characterized
by
different
microclimatic
conditions,
soil
texture,
soil
thawing
rate,
soil
moisture
and
thermokarst
processes
(Dolman
et
al.,
2004).
The
thickness
of
the
thin
active
layer
that
separates
the
forest
from
the
permafrost
is
highly
susceptible
to
hot-dry
or
dry/wet
climate
alike
(Sugimoto
et
al.,
2003;
Iwahana
et
al.,
2005;
Lopez
et
al.,
2007b),
which
is
important
considering
rainfall
extremes
predicted
as
a
result
of
global
warming
(Karl
et
al.,
1995).
Despite
the
growing
interest
in
this
region,
flux
measurements
on
grassland
ecosystems
have
been
limited
to
greenhouse
gas
(GHG)
flux
measurements
by
portable
chambers
from
differ-
ent
vegetation
grassland
belts
within
the
grassland
ecosystem
(Morishita
et
al.,
2004;
Takakai
et
al.,
2008).
The
lack
of
carbon
and
water
flux
measurements
in
this
ecosystem
has
under-
mined
our
understanding
of
this
ice-rich
permafrost
region
in
Central
Yakutia.
Models
of
water
and
carbon
fluxes
in
eastern
Siberia
need
to
consider
the
role
that
grasslands
play
in
the
short-
and
long-term
if
they
intend
to
predict
reliably
the
consequences
of
climate
change
in
this
region.
There
are
still
many
factors
controlling
carbon
turnover
in
high
latitudes
that
are
not
well
understood
(Hobbie
et
al.,
2000).
The
aims
of
this
study
are:
(1)
to
determine
the
environmental
factors
controlling
carbon
and
water
vapor
fluxes
from
forest
and
alas
(grassland)
and
(2)
to
assess
the
differences
of
carbon
and
water
vapor
fluxes
from
forest
and
alas
ecosystems
at
the
regional
level
in
Central
Yakutia.
2.
Material
and
methods
2.1.
Study
site
Neleger
Experimental
Station
is
located
30
km
north-north-
west
of
the
city
of
Yakutsk
(62°05'N,
129°45')
and
is
part
of
a
network
of
stations
of
the
Yakutian
Permafrost
Institute,
Russian
Academy
of
Science,
in
eastern
Siberia
(Fig.
1).
Mean
annual
air
temperature
is
-10
to
-11
°C;
amplitude
of
monthly
temperatures
is
about
62
°C.
Annual
snow
cover
ranges
from
30
to
40
cm,
with
a
maximum
of
60
cm.
Ice
wedges
(ice
content
of
more
than
90%)
are
distributed
in
the
permafrost
with
the
upper
edge
found
at
depth
of
1.5
m
from
the
soil
surface.
Precipitation
during
the
snow-free
growing
season
is
approxi-
mately
110
mm,
which
is
about
half
the
annual
precipitation,
whereas
the
corresponding
potential
evaporation
rate
is
370
mm
(Muller,
1982).
Soils
in
the
region
are
classified
as
Gelisols
and
their
texture
is
silty-clay-loam
(SiCL)
to
silty-clay
(sic).
Two
sites
were
chosen
for
flux
and
micrometeorological
measurements:
forest
and
alas.
The
forest
is
dominated
by
200-year-old
larch
(Larix
cajanderi)
trees
with
average
height
of
8.6
m.
The
maximum
tree
height
is
21
m
and
average
tree
density
is
2100
tree
ha
-1
.
The
forest
floor
is
covered
with
shrubs,
and
moss
dominated
by
Vaccinium
vitis
idaea.
At
the
alas
site,
several
belts
of
concentric
vegetation
are
distributed
from
the
edge
of
the
alas
to
the
pond
at
the
center
of
the
depression.
The
belt
closer
to
the
pond
is
dominated
by
Scholocloa
festucacea.
The
next
vegetation
belt
with
high
soil
water
content
is
dominated
by
Carex
vesicata
and
other
minor
plants.
The
following
belt
is
an
intermediate
zone
between
the
1970
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
130°E
Neleger
Yakutsk
o`f
I
F
site
site
0
100
200
100
400
rn
Dry
grassland
Wet
grassland
Pond
Ping
Fig.
1
-
Location
of
the
study
site
and
position
of
the
forest
site
(F-site,
open
circle)
and
the
alas
site
(A-site,
filled
circle).
Different
area
colors
represent
the
dry
and
wet
vegetation
with
the
lake
in
the
middle.
wet
and
dry
soil
belt,
which
is
mainly
composed
by
Carex
angarae.
Finally
the
outer
belt,
which
is
characterized
by
dry
soil
conditions,
is
dominated
by
Elytrigia
repens.
Vegetation
in
the
Pingo,
a
small
hill
(almost
3
m
high),
at
the
edge
of
the
lake
in
the
alas,
is
dominated
by
Carex
duriuscula
and
Elytrigia
repens.
The
lower
part
of
the
alas
site
is
approximately
3
m
below
the
relative
height
of
the
forest
site.
Soil
moisture
and
chemical
composition
of
the
soil
in
the
forest
and
alas
site
has
been
thoroughly
described
in
Lopez
et
al.
(2007a).
The
area
of
the
studied
alas
is
approximately
32
ha.
2.2.
Measurements
Meteorological
and
flux
measurements
were
conducted
from
23
May
to
31
August
during
the
growing
season
of
2006.
Meteorological
measurements
were
conducted
on
a
21
m
high
tower
at
the
forest
site
as
follows:
wind
speed
at
21
m
(3-cups
anemometer,
R.M.
Young,
31012,
USA);
net
radiation
(R
a
)
at
21
m
(Kipp
&
Zonen,
CNR1,
Netherlands);
air
temperature
(T
a
)
and
relative
humidity
(RH)
at
0.5,
2.0,
8.0,
14.0,
18.0
and
21.0
m
height
(Vaisala
HMP45A,
Finland).
For
the
eddy
flux
covariance
an
ultrasonic
anemometer
(Gill
Solnet,
R3,
England)
and
open
path
CO
2
/H
2
0
gas
analyzer
(Licor
7500,
USA)
were
installed
at
the
top
of
the
tower.
In
the
alas
site
the
same
instrumentation
was
installed
for
eddy
flux
covariance
and
meteorological
measurements:
R.
(Kipp
&
Zonen,
CNR1,
Netherlands);
T
a
and
RH
(Vaisala
HMP45C,
Finland);
rainfall
(tipping
bucket,
TR-
525M,
Texas
Electronics,
USA)
and
wind
speed
(3-cups
anemometer,
R.M.
Young,
31012,
USA)
were
measured
at
a
2
m
high
tower.
Ground
heat
flux
(G)
was
obtained
by
the
average
of
outputs
from
three
heat
flux
plates
(REBS,
model
HFP01).
Simultaneously,
lake
water
level
(HM-500
series,
Hi-
net,
Japan)
in
the
alas
site
was
measured.
Sensible
heat
(H),
water
vapor
(LE)
and
carbon
dioxide
fluxes
(F
e
)
were
calculated
for
30
min
intervals.
A
linear
trend
removal
for
CO
2
concentration
and
a
time
lag
correction
using
maximum
covariance
were
applied
before
calculating
fluxes.
The
WPL
correction
for
air
density
fluctuation
was
applied
(Webb
et
al.,
1980).
Stationarity
of
the
flux
was
examined
by
comparing
covariance
over
a
30
min
period
and
averaging
of
the
covariance
for
six
5
min
sub-periods,
eliminating
those
for
which
when
the
relative
difference
exceeded
40%.
To
examine
the
effective
fetch,
footprint
distance
of
the
surface
fluxes
from
the
towers
was
analyzed
and
the
fluxes
with
the
windward
fetch
shorter
than
80%
of
footprint
distance
were
eliminated
(Korman
and
Meixner,
2001).
Data
gaps
during
the
whole
study
period,
due
to
missing
and
rejected
data
were
about
63%
for
CO
2
,
58%
for
water
vapor
and
53%
for
sensible
heat
flux
in
the
alas
site.
In
the
forest
site
following
the
same
rejection
criteria
and
missing
data,
the
gaps
were
48%
for
CO2,
44%
for
water
vapor
and
51%
for
sensible
heat
flux
respec-
tively.
In
both
cases
the
gaps
occurred
generally
in
the
evening
with
approximately
70%
in
the
alas
site
and
60%
in
the
forest
site.
Friction
velocity
was
applied
as
a
filter
(<0.1),
for
night
time
data.
The
method
used
to
fill
the
gaps
was
multiple
imputation
(Hui
et
al.,
2004).
The
reliability
of
the
eddy
flux
covariance
measurements
was
examined
by
the
energy
balance
closure
(Wilson
et
al.,
2002)
at
each
of
the
sites.
The
linear
regression
between
half-hourly
values
of
the
available
energy
(R.
-
G)
and
the
eddy
fluxes
(H
+
LE)
in
the
alas
site
gave
an
intercept,
slope
and
coefficient
of
determination
(r
2
)
of
22.9
W
m',
0.72
and
0.81
(P
<
0.0001)
respectively.
The
values
for
the
forest
site
energy
balance
closure
were
14.8
W
m
-2
,
0.84
and
0.87
(P
<
0.0001)
respectively.
These
results
suggest
that
the
eddy
flux
covariance
underestimated
H
+
LE
by
28%
and
16%
in
the
alas
and
forest
site
respectively.
As
can
be
observed
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
1971
from
these
values
the
energy
balance
closure
is
not
perfect,
especially
for
the
alas
site
and
one
of
the
main
reasons
is
that
the
value
for
G
was
taken
from
a
5-year
clear-cut
site
next
to
the
alas
(150
m
away
from
the
alas
tower).
The
characteristics
of
the
alas
site,
with
different
grass
type
and
biomass
makes
it
difficult
to
obtain
a
single
G
value.
Canopy
conductance
was
derived
from
the
Penman-
Monteith
equation:
(a)
_s
r
VPD
3.0
-
2.5
-
2.0
-
1.5
1.0
0.5
5,0
A(R
a
-
G)
+
pC
p
VPD
g
a
LE
-
[A
+
Y[
1
+
(9a
+
9a)]]
(b)
-T
a
450
300
150
30
25
20
where
LE
is
the
latent
heat
(W
m'),
A
the
rate
of
change
of
saturation
vapor
pressure
(Pa
C
-1
),
R
a
the
net
radiation
above
the
stand
(W
m
-2
),
G
the
heat
flux
in
the
soil
(W
m'),
p
the
density
of
dry
air
at
constant
pressure
U
kg
-1
C
-1
),
VPD
the
vapor
pressure
deficit
(Pa),
g
a
the
aerodynamic
conductance
(m
s
-1
),
g
c
the
canopy
conductance
(m
s
-1
)
and
y
is
the
psy-
chrometric
constant
(Pa
C
-1
).
Aerodynamic
conductance
(ga,
m
s
-1
)
was
calculated
as
k
2
u
g
a
{ln[(z
-
d)/z
0
]}
,
'
i
(c)
-
4.0
-
3.0
1
3
2.0
-
1.0
ar,
15
10
50
40
30
20
10
The
displacement
height
(d)
was
set
as
0.67h
and
the
rough-
ness
length
(z
0
)
as
0.1h,
where
h
is
stand
height,
k
von
Kar-
man's
constant
(-:-.0.40)
and
u
is
wind
speed
at
height
z
above
the
canopy
(Brutsaert,
1982).
In
order
to
estimate
the
area
of
each
of
the
vegetation
belts
(dry,
wet)
and
the
lake
area,
handheld
global
positioning
system
(GPS)
equipment
(eTrex,
Garmin,
USA)
was
used.
Coordinate
information
obtained
by
the
GPS
was
used
as
input
data
in
the
GIS
software
Arc
GIS
9.0
(ESRI
Japan,
2004).
3.
Results
3.1.
Environmental
conditions
Environmental
conditions
at
the
study
site
in
2006
are
shown
in
Fig.
2.
Air
temperatures
were
higher
in
the
second
half
of
July
and
the
beginning
of
August
in
2006.
The
first
half
of
July
was
a
period
of
inclement
weather,
giving
rise
to
a
two
apparent
peaks
of
solar
radiation
S
r
,
T
a
and
VPD
during
the
study
period
in
June
and
late
July.
In
August
all
these
variables
were
at
their
lowest
point
because
of
frequent
rain
events.
Wind
speed
ranged
between
1.0
and
4.0
m
s
-1
and
was
on
average
2.5
m
s
-1
.
As
is
characteristic
of
this
region
precipita-
tion
was
distributed
unevenly
with
11.4%,
13.5
%
and
74.4
%
in
June,
July
and
August
respectively.
During
the
period
May-
August,
the
total
precipitation
was
214.4
mm.
During
May,
June
and
July
the
water
level
of
the
lake
in
the
alas
ecosystem
decreased
at
a
rate
of
0.5
cm
d
-1
.
This
decrease
is
a
surrogate
of
lake
water
evaporation
under
the
dry
conditions
in
these
months.
The
water
level
increased
when
extreme
precipita-
tion
events
occurred
on
6,
14
and
21
August
(17,
34
and
41
mm,
respectively).
As
expected
the
area
of
the
lake
shrank
as
the
water
level
decreased.
Surface
runoff
from
the
forest
to
the
alas
ecosystem
in
summer
is
unlikely
considering
that
by
this
time
the
soil
thawing
depth
in
the
dry
vegetation
grassland
Lit
May
June
Fig.
2
-
Environmental
conditions
during
the
period
May
23rd-August
31st.
(a)
Solar
radiation,
Sr
and
vapor
pressure
deficit,
VPD;
(b)
air
temperature,
Ta
and
wind
speed,
u;
(c)
precipitation,
rain
and
alas
lake
water
level.
(which
lies
between
the
forest
and
the
lake)
was
about
130-
150
cm
and
that
soil
water
infiltration
was
high
(Lopez
et
al.,
2007a).
During
this
period,
leaf
area
increases
in
the
forest
and
alas
vegetation
were
observed,
without
any
apparent
drought
stress
due
to
low
precipitation
in
May
and
June.
3.2.
Diurnal
fluxes
from
alas
and
forest
ecosystems
The
influence
of
environmental
variables
on
F
c
and
LE
was
examined
on
three
continuous
days
15,
16
and
17
July.
These
days
were
chosen
because
of
their
cloudless
conditions
and
because
the
data
gaps
were
minimal,
especially
during
the
daylight
hours
(Fig.
3a).
During
these
three
days
the
pre-
dominant
wind
was
from
the
NW-NE,
which
is
the
direction
of
the
fluxes
over
the
lake.
LE
in
the
forest
started
decreasing
after
VPD
increased
more
than
1.2-1.5
kPa
and
the
LE
in
the
alas
reached
the
maximum
value
together
with
the
maximum
S
r
.
LE
in
the
forest
is
slightly
larger
than
LE
in
the
alas
at
midday
but
in
general
the
difference
was
not
significant.
LE
fluxes
in
the
forest
and
in
the
alas
started
and
ended
at
the
same
time
in
the
morning
and
late
afternoon.
In
contrast,
carbon
uptake
in
the
forest
and
in
the
alas
showed
a
larger
difference
not
only
at
midday
but
all
throughout
the
three
days
(Fig.
3b).
Carbon
uptake
in
the
forest
started
and
peaked
earlier
in
the
morning
compared
to
the
carbon
uptake
in
the
alas.
The
midday
decline
of
carbon
uptake
observed
in
the
forest
was
not
observed
in
the
alas
(Fig.
3c).
0.0
0.6
0.4
0,2
00
(LI
)
eV
el
SeiV
Rain
-
Watley
July
August
1972
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
(a)
Jul-16
Jul-17
2.5
2.0
1.5
tzp
-
1.0
-
0
-
0.5
0.0
(b)
-
0-
alas
-
0-
forest
A
(c)
6
12
18
0
6
12
18
0
6
12
18
Time
(hour)
Fig.
3
-
Diurnal
environmental
conditions
on
days
July
15,
16
and
17
(upper
row
of
graphs);
their
respective
diurnal
water
vapor
flux
(LE)
(mid
row)
and
the
CO
2
(lower
row)
flux
at
the
alas
(open
circles)
and
forest
(filled
circles)
sites.
-J
0
e
l
4
E
U-
250
200
150
100
50
0
0.3
02
0.1
0.0
-0.1
-0.2
-0.3
0
800
600
E
400
200
When
air
temperature
increased in
the
morning,
VPD
and
S
r
drive
the
free
opening
of
stomata
of
wet
grassland
vegetation
and
transpiration
increased
accordingly.
This
is
accompanied
by
transpiration
from
vegetation
in
the
dry
grassland
as
well
as
soil
and
lake
evaporation.
Together,
these
account
for
the
total
water
loss
from
the
alas
ecosystem,
which
nearly
equals
the
water
loss
from
the
forest.
On
the
other
hand,
carbon
uptake
peaks
early
in
the
morning
but
at
around
10:00
h,
it
comes
to
a
halt
and
instead
a
plateau
is
observed
that
continued
until
late
afternoon
when
S
r
started
declining.
The
difference
between
the
diurnal
courses
of
carbon
uptake
in
the
alas
appears
to
be
related
to
hetero-
trophic respiration
from
the
dry
grassland,
which
peaks
late
in
the
morning
and
continues
until
late
afternoon
(Takakai
et
al.,
2006),
This
respiration
offsets
the
GEP
of
the
wet
grassland
and
consequently
lowers
the
NEE
of
the
alas
ecosystem.
This
characteristic
has
the
potential
to
influence
significantly
its
annual
NEE,
because
of
the
dry/wet
cycles
in
eastern
Siberia
(Shender
et
al.,
1999)
that
determine
the
annual
fluctuation
of
dry/wet
grassland
area
within
the
alas
ecosystem.
Several
days
during
the
study
period
were
selected
for
the
alas
site
to
corroborate
the
effect
of
the
prevailing
wind
in
the
fluxes
from
this
ecosystem.
When
winds
came
from
the
SW-
SE
the
surface
was
only
grassland
but
when
it
came
from
the
NW-NE
the
surface
included
the
alas
lake.
The
results
showed
that
LE
fluxes
were
generally
higher
when
the
water
source
was
the
lake.
On
the
other
hand,
carbon
uptake
difference
due
to
flux
source
direction
was
not
clear.
3.3.
Daily
evaporation
and
net
ecosystem
exchange
at
the
forest
and
alas
ecosystems
At
the
end
of
May
and
until
mid-June,
alas
E
was
on
average
18%
higher
than
E
at
the
forest
but
when
the
tree
canopy
was
fully
foliated,
forest
E
became
approximately
23%
higher
for
the
rest
of
the
season
(Fig.
4a).
In
mid-July
forest
E
decreased,
due
to
low
incoming
radiation,
high
humidity
and
low
temperatures.
The
evaporative
demand
was
the
same
for
both
sites
but
forest
E
reached
a
maximum
of
2.3
mm
d
-1
while
the
maximum
in
the
alas
site
was
1.9
mm
c1
-1
.
Mean
E
rates
in
the
forest
was
1.4
mm
d
-1
and
in
the
alas
was
1.2
mm
c1
-1
.
The
values
in
the
forest
are
similar
to
values
reported
in
other
studies
in
larch
in
eastern
Siberia
(Baldocchi
et
al.,
2000;
Ohta
et
al.,
2001;
Dolman
et
al.,
2004).
The
values
reported
in
this
study
for
the
alas
are
difficult
to
compare
with
previous
studies
in
this
region
because
to
our
knowledge
there
is
no
study
of
hydrological
measurements
from
the
total
area
or
from
each
of
the
grass
belts
or
the
lake
of
an
alas
ecosystem.
According
to
the
review
by
Kelliher
et
al.
(1993),
maximum
daily
E
from
grassland
in
temperate
zones
ranged
from
4.1
to
6.2
mm
c1
-1
.
These
maximum
values
are
2-25-fold
higher
than
the
value
found
in
the
boreal
grassland
of
our
study.
Lower
values
have
been
reported
for
a
tussock
grassland
in
New
Zealand
under
dry
conditions
where
the
maximum
E
was
3.8
mm
d
-1
and
the
mean
value
was
1.1
mm
C1
-1
(Hunt
et
al.,
2002),
while
in
a
grassland
in
the
Loess
Plateau
in
China
the
mean
value
was
1.0
mm
C1
-1
(Kimura
et
al.,
2006).
-a
E
E
Lu
2.5
2.0
1.5
1.0
0.5
12
10
E
E
-10
2
0
-4
-6
-8
NEE
(g
Cm
-
2
c
1
-
1
)
10
8
6
4
2
-10
NEE
(g
C
m
-
2d
-
1)
0/2
0
-2
-4
-6
-8
0
0
o
o
s
forest
o
alas
r•T
re
,
=
0
79
(a)
ca
g
CO
CD
021
0
00
5
alas
=
0.82
(b)
0
0
0
=
0.24
0
0
:it•
0
0
o
coo
0
CD
D
.
0
0
0
11
0
Soo
1111
r'
for
=
0.56
f
s•
i
Sig
2P
e
O
O
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008)
1968-1977
1973
(a)
-0-
alas
Ar-
forest
_
(b)
May
June
July
August
Fig.
4
-
(a)
Daily
evapotranspiration
(E)
(b)
canopy
conductance
(g
c
);
(c)
NEE
in
the
alas
(open
circles)
and
the
forest
(filled
circles)
during
the
period
May
23rd-August
31st.
The
mean
forest
canopy
conductance
(g
c
)
during
June
and
July
was
3.6
mm
s
-1
(Fig.
4b).
For
the
larch
forest
in
this
region
Dolman
et
al.
(2004)
reported
values
of
2-3
mm
s
-1
whereas
Ohta
et
al.
(2001)
reported
a
mean
value
of
about
4
mm
s
-1
(canopy
resistance,
r
c
=
250
s
m
-1
)
for
June
and
July
with
maximum
values
found
in
August
in
the
same
larch
forest
as
the
study
mentioned
above.
The
canopy
conductance
of
the
alas
had
lower
values
to
those
found
in
the
forest
but
like
the
forest
g
c
,
it
was
controlled
by
VPD
(Fig.
5a).
The
differences
in
LAI
in
the
forest
and
in
the
alas
are
assumed
to
be
one
of
the
main
reasons
for
the
difference
observed
between
sites.
To
be
precise,
g
c
in
the
alas
site
is
not
from
one
surface
alone
but
a
composition
of
different
grass
surfaces
under
different
soil
moisture
conditions
(Lopez
et
al.,
2007a)
and
the
lake.
NEE
in
the
alas
is
less
controlled
by
VPD,
reflecting
the
multiple
other
factors
that
are
in
play
and
which
are
a
consequence
of
phenological
differences
in
grasses
and
the
moisture
condi-
tions
in
each
of
the
belts
(Fig.
4c).
All
these
differences
are
perceived
and
integrated
by
the
eddy
covariance
system.
On
the
other
hand,
a
stronger
relationship
between
g
c
and
VPD
in
the
forest
is
paired
with
a
strong
relationship
between
VPD
and
NEE
in
the
forest
(Fig.
5).
NEE
in
the
forest
site
was
more
negative
(meaning
more
carbon
uptake
by
the
ecosystem)
than
the
alas
NEE
during
the
entire
growing
season.
Carbon
sequestration
in
the
alas
00
0.5
1.0
1,5
2.0
2,5
VPD
(kPa)
Fig.
5
-
(a)
Relationship
between
the
canopy
conductance
(g
c
)
and
the
vapor
pressure
deficit
(VPD)
for
the
alas
(open
circles)
and
the
forest
(filled
circles);
(b)
relationship
between
NEE
and
VPD
for
the
alas
and
forest
(symbols
as
above).
This
data
correspond
to
fine
weather
days.
ecosystem
ceased
approximately
at
the
end
of
July
more
than
one
month
before
it
ceased
at
the
forest.
The
maximum
daily
NEE
from
the
forest
was
-9.0
gC
m
-2
d
-1
and
-4.0
gC
m
-2
from
the
alas
ecosystem.
The
mean
daily
NEE
was
-4.5
gC
m
-2
s
-1
in
the
forest
and
-1.2
gC
m
-2
s
-1
in
the
alas
ecosystem.
In
previous
studies,
a
similar
daily
value
was
found
for
a
tussock
grassland
(Hunt
et
al.,
2002).
3.4.
Seasonal
energy,
water
and
carbon
balance
in
the
alas
and
forest
site
In
May
when
leaves
were
not
fully
grown,
alas
E
was
higher
than
forest
E.
In
June
and
July,
when
the
leaves
were
fully
grown,
forest
E
was
17%
higher
than
in
the
alas
ecosystem
(Table
1).
Evapotranspiration
from
May
to
August
2006
was
136
mm
and
118
mm
in
the
forest
and
alas
respectively.
The
ratio
of
LE/12,
1
has
a
mean
of
0.30
for
the
forest
and
0.36
for
the
alas,
consistent
with
mean
evapotranspiration
rates.
This
shows
the
difference
in
available
energy
among
vegetation
types
because
forests
reflect
less
shortwave
radiation
than
grasslands.
Lake
water
level
decreased
3.3,
14.8
and
11.8
cm
in
May,
June
and
July
while
in
August
the
lake
water
level
increased.
These
changes
in
the
lake
water
level
were
strongly
related
to
E
(8.5,
44.4,
40.7
and
24.8
mm)
and
precipitation
(1.4,
24.4,
29.0
and
159.6
mm)
in
May,
June,
July
and
August
1974
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
Table
1-
Monthly
values
of
sensible
(H),
latent
heat
flux
(LE),
Bowen
ratio
(Bo),
ratio
of
LE/R„,
evapotranspiration
(E)
and
NEE
in
the
forest
and
in
the
alas
site
for
study
period
of
2006.
H
(W
m
-2
)
LE
(W
m
-2
)
Bo
LE/R
a
E
a
(mm)
F
c
(tC
ha
-1
)
Forest
Alas
Forest
Alas
Forest
Alas
Forest
Alas
Forest
Alas
Forest
Alas
May
23-31
147.8
57.0
36.4
35.8
4.1
1.6
0.15
0.22
6.7
8.5
-0.25 -0.05
June
01-30
158.6
58.5
81.8
70.9
1.9
0.8
0.26
0.37
48.5
44.4
-1.66
-0.77
July
01-31
125.2
62.2
80.3
69.4
1.6
0.9
0.30
0.33
49.4
40.7
-1.08
-0.43
August
01-31
72.4
45.9
74.6
55.0
1.0
0.8
0.45
0.31
31.6
24.8
-0.45
-0.14
Total
136.2
118.4
-3.44
-1.38
respectively.
The
results
indicate
that
in
any
given
day,
E
was
similar
in
the
forest
and
alas
ecosystems,
but
this
difference
(Eforest
-
Ealas)
became
larger
when
the
flux
origin
was
from
the
grassland
rather
than
from
the
lake.
The
maximum
NEE
was
reported
on
30
June
(4.0
gC
m'
d
-1
)
which
appears
to
be
the
turning
point
where
the
NEE
starts
decreasing
until
the
end
of
July
when
the
alas
ecosystem
becomes
a
source
after
being
a
sink
for
approxi-
mately
two
months
(June
July).
The
NEE
in
the
forest
ecosystem
was
more
than
twice
(-3.44
tC
ha
-1
)
that
of
NEE
in
the
alas
ecosystem
(-1.38
tC
ha
-1
)
during
the
period
of
study.
June
and
July
were
the
most
active
months
with
forest
uptaking
four
times
more
CO
2
than
the
alas
ecosystem,
at
the
same
time
alas
E
for
the
period
considered
was
only
13%
lower
than
forest
E.
A
rainless
period
that
extended
from
14
July
to
2
August
produced
a
decline
in
evapotranspiration
and
carbon
uptake.
This
decline
is
commonly
associated
with
the
continuous
depletion
of
soil
moisture
in
the
upper
soil
layers
(Lopez
et
al.,
2007b)
where
80%
of
the
root
system
is
distributed
(Kuwada
et
al.,
2002).
This
shows
the
importance
of
soil
thawing
rate;
while
the
rainless
span
was
similarly
long
at
the
beginning
of
the
growing
season,
the
shallow
thawed
layer
stored
more
of
the
little
water
available
for
root
water
uptake,
allowing
the
increase
of
LE
and
NEE.
On
mid-July
the
soil
thawing
depth
is
around
100
cm
and
precipitated
water
does
not
remain
in
the
upper
soil
layers
but
moves
downward
to
deeper
layers
(Iwahana
et
al.,
2005).
NEE
in
the
forest
was
larger
than
that
reported
in
previous
studies
for
the
same
region
(Dolman
et
al.,
2004;
Machimura
et
al.,
2005).
In
2001,
Dolman
et
al.
(2004)
reported
an
annual
forest
carbon
uptake
of
-2.3
tC
ha
-1
and
for
the
same
year
Machimura
et
al.
(2005)
reported
a
growing
season
value
of
-1.1
tC
ha
-1
,
which
is
one-third
of
the
NEE
reported
in
the
forest
(-3.4
tC
ha
-1
)
in
this
study.
It
is
possible
that
our
nighttime
value
was
underestimated
by
the
gap
filling
method
and
for
that
reason
the
NEE
shows
more
negative
values
(more
CO
2
uptake).
The
NEE
in
the
alas
had
a
rather
conservative
value
compared
to
NEE
from
grasslands
in
previous
studies,
-1.93
to
-2.58
tC
ha
-1
depending
on
climate
and
soil
moisture
conditions
(Jaksic
et
al.,
2006)
and
it
is
even
lower
than
the
value
of
-2.4
tC
ha
-1
found
for
nine
grasslands
across
Europe
(Soussana
et
al.,
2007).
This
latter
value
is
for
a
year
total,
that
is,
considering
the
winter
period
in
which
it
is
expected
to
behave
as
a
source
of
CO
2
.
If
this
is
not
included,
as
in
our
results,
the
carbon
sequestration
reported
in
this
latter
study
would
be
even
higher.
4.
Discussion
This
study
was
conducted
during
the
growing
season
of
2006,
with
a
total
rainfall
of
214.4
mm
while
rainfall
during
the
growing
season
of
2005
was
208.9
mm
(reported
for
the
same
site,
but
not
shown).
This
means
that
moisture
accumulated
in
the
soil
profile
was
high
at
the
beginning
of
the
growing
season
in
2006.
Soil
moisture
storage
gives
rise
to
a
state
in
which
the
previous
year's
precipitation
is
held,
after
freezing
in
winter,
and
gradually
released
during
the
growing
season
as
the
soil
thaws
(Sugimoto
et
al.,
2003;
Iwahana
et
al.,
2005;
Lopez
et
al.,
2007b).
This
mechanism
was
observed
in
the
forest
as
well
as
in
the
alas.
The
difference
is
that
in
the
alas
there
are
different
belts
of
vegetation
and
soil
moisture
conditions,
which
influence
the
soil
thawing
rate
(Lopez
et
al.,
2007a).
The
rule
is
that
dry
soils
thaw
faster
than
wet
soils
because
of
the
thermal
conductivity
and
latent
heat
of
the
soil
moisture
(Romanovsky
et
al.,
1997a).
The
maximum
soil
thawing
depth
in
the
forest,
dry
grassland
and
wet
grassland
is
1.1,
1.8
and
1.2
m
respectively
(Lopez
et
al.,
2007a).
In
this
study,
precipitation
in
May
and
June
combined
was
25.8
mm
in
total
but
E
in
the
same
period
was
52.9
and
55.2
mm
in
the
forest
and
alas
respectively.
Water
fluxes
and
E
in
the
forest
are
similar
to
studies
conducted
in
the
same
region
for
larch
stands
(Kelliher
et
al.,
1997;
Ohta
et
al.,
2001;
Dolman
et
al.,
2004).
Drastic
inter-annual
climatic
variation
do
not
appear
to
affect
seasonal
forest
water
balance
as
it
has
been
pointed
out
by
Sugimoto
et
al.
(2003)
and
Dolman
et
al.
(2004)
nevertheless
its
effect
on
inter-annual
variation
of
carbon
fluxes
appeared
to
be
significant
(Dolman
et
al.,
2004;
Machimura
et
al.,
2005).
This
phenomenon
is
the
result
of
VPD
control
on
forest
evapotranspiration
(Dolman
et
al.,
2004);
slow
soil
thawing
rate
(Lopez
et
al.,
2007b)
and
soil
moisture
stored
from
the
previous
year
growing
season
(Sugimoto
et
al.,
2003).
As
mentioned
throughout
this
study,
forest
E
has
been
measured
before
in
this
region
but
not
alas
E
and
since
our
data
includes
only
one
growing
season,
it
is
impossible
to
evaluate
the
same
response
from
this
ecosystem,
but
considering
that
in
a
dry
year
the
lake
area
is
smaller,
it
is
expected
that
E
will
be
smaller
too.
The
water
vapor
flux
data
analysis
in
the
alas
indicates
that
the
growing
season
alas
E
is
merely
13%
lower
than
the
forest
E
despite
the
obvious
difference
of
this
two
sites
in
terms
of
canopy
architecture.
This
result
coincides
with
comparisons
between
coniferous
forest
and
grasslands
in
a
review
study
(Kelliher
et
al.,
1993)
but
the
peculiar
characteristics
of
the
alas
make
it
difficult
to
focus
on
one
particular
source
to
explain
E.
In
this
site,
it
is
more
precise
to
divide
its
contribution
among
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008)
1968-1977
1975
the
lake,
the
wet
grassland
and
the
dry
grassland.
Lake
water
level
decrease,
a
qualitative
surrogate
for
water
evaporation,
indicates
larger
water
loss
during
rainless
periods.
Further-
more,
wet
grassland
(around
27%
of
the
total
area)
is
thought
to
be
another
significant
source
for
E
because
of
its
denser
aboveground
vegetation
(560
g
biomass
m
-2
)
compared
to
the
dry
grassland
(around
70%
of
the
total
alas)
where
the
biomass
in
on
average
223
g
biomass
m'
(Takakai
et
al.,
2008).
Soil
moisture
ranges
from
35
to
50%
in
the
wet
grassland
and
from
18
to
35%
in
the
dry
grassland
during
the
growing
season
(Lopez
et
al.,
2007a).
The
mechanisms
and
partition
of
forest
E
has
been
shown
in
several
studies
(Arneth
et
al.,
1996;
Ohta
et
al.,
2001;
Lopez
et
al.,
2007b).
The
partition
was
about
50-50%
between
the
understory
and
the
tree
canopy
for
each
of
the
sources
mentioned
above.
Actually
in
Lopez
et
al.
(2007b),
the
partition
of
evapotranspiration
shows
an
apparent
inter-
annual
variation
from
50-50%
when
it
was
a
rainy
growing
season
to
40-60%
(understory-tree
canopy)
during
a
dry
growing
season.
Canopy
conductance
from
the
forest
and
the
alas
site
showed
different
values
but
similar
tendencies
during
the
months
of
June
and
July.
Forest
g
c
has
the
dual
source
of
tree
canopy
and
understorey
vegetation,
while
in
the
alas
ecosystem
the
origin
has
multiple
sources
that
would
need
to
be
studied
more
in
detail.
Mean
values
differ
as
expected
from
the
different
canopy
architecture
of
this
two
ecosystems.
It
is
already
common
knowledge
that
VPD
controls
g
c
in
coniferous
forest
but
as
shown
in
this
study
it
also
exerts
a
significant
control
on
g
c
in
the
alas
ecosystem.
NEE
reported
in
this
study
was
higher
than
values
found
by
previous
studies
in
this
region
in
the
same
larch
forest
in
Central
Yakutia
(Dolman
et
al.,
2004;
Machimura
et
al.,
2005).
In
the
first
study
gap
filling
was
not
applied
and
in
the
second
study
calculation
of
seasonal
NEE
was
based
on
14
days
mean
diurnal
variation
that
could
have
led
to
significant
under-
estimation
because
of
the
non-linear
dependence
of
NEE
on
environmental
variables.
Another
difference
with
the
second
study
is
that
they
used
the
closed-path
system
(Li-6262).
Night
time
correction
using
friction
velocity
was
applied
in
both
cases
(<0.4
and
0.1
m
s
-1
,
respectively).
In
our
case,
friction
velocity
(<0.1)
was
applied
as
a
filter,
for
night
time
data.
It
has
been
observed
that
years
with
high
rainfall
regimes
increase
the
leaf
area
for
the
next
growing
season
(Saito
et
al.,
2005)
and
this
can
explain
the
higher
NEE
in
2006
compared
with
previous
studies.
This
same
phenomenon
has
not
been
observed
in
the
alas
and
could
be
one
of
the
reasons
for
the
large
difference
between
NEE
in
the
forest
and
the
alas
ecosystem
in
this
study.
In
the
alas
ecosystem
flooding
is
the
mechanism
that
accumulates
carbon
in
wet
years,
which
is
subsequently
released
in
dry
years.
Soil
respiration
in
the
alas
increases
when
the
area
of
vegetation
inundated
decreases,
and
the
size
of
the
lake
shrinks
(Takakai
et
al.,
2006).
A
drier
and
warmer
climate
will
increase
carbon
emissions
from
the
grassland
because
of
aerobic
decomposition
of
organic
matter
stored
in
the
waterlogged
soils
(in
wet
years).
Thus,
it
appears
that
NEE
is
largely
influenced
by
ecosystem
respiration
(ER)
while
in
the
forest
GEP
(Gross
ecosystem
production)
drives
NEE
(Dolman
et
al.,
2004).
In
(Hirano
et
al.,
2006),
the
annual
course
of
ER
in
the
larch
forest
peaks
from
the
end
of
July
to
August
in
the
year
2004
and
2005
(6.0
and
8.0
gC
m' d'
respectively).
This
increase
coincided
with
the
decrease
of
NEE
observed
from
late
July
in
this
study.
Nonetheless,
NEE
throughout
the
growing
season
is
negative,
showing
the
predominance
of
GEP
over
ER.
This
is
one
example
of
the
importance
of
hydrology
on
carbon
uptake.
If
grasslands
expand,
water
vapor
flux
in
the
forest
and
the
alas
ecosystem
will
be
relatively
similar,
provided
it
is
a
rainy
year
but
based
on
the
results
found
in
this
study
we
assume
that
in
a
dry
year
evapotranspiration
from
the
forest
will
be
higher
than
from
the
alas
because
of
deeper
root
system
and
slower
thawing
rate.
Moreover,
because
of
smaller
lakes,
increased
soil
respiration
will
decrease
NEE,
making
the
difference
in
carbon
uptake
between
the
larch
forest
and
the
alas
larger.
Furthermore,
dry/wet
or
warm/cool
years
occur
in
cycles
in
eastern
Siberia
(Shender
et
al.,
1999).
In
the last
four
years
annual
precipita-
tion
average
(2003-2006,
286
mm)
has
resulted
in
larger
areas
of
inundated
vegetation
in
the
alas
ecosystems
in
contrast
to
the
drier
average
of
the
previous
three
years
(1999-2002,
153
mm).
All
this
must
be
put
in
long-term
context
though,
as
in
the
end
it
will
be
the
hydrology
(and
higher
air
tempera-
tures)
of
this
region
that
will
drive
carbon
uptake.
It
is
also
necessary
to
consider
snow
melting
in
spring
because
high
snow
depth
can
form
a
lake
so
big
in
the
depression
(surface
runoff)
that
the
entire
alas
might
become
a
single
lake
(as
was
observed
in
2007).
Based
on
the
total
area
of
alas
(grassland)
in
lowland
Central
Yakutia
of
440
thousand
ha,
NEE
from
alas
ecosys-
tems
was
roughly
estimated
as
0.6
x
10
-3
Pg.
Meanwhile
NEE
for
a
similar
forested
area
is
1.5
x
10
-3
Pg.
Most
of
models
in
eastern
Siberia
ignore
the
presence
of
alases
and
just
considered
all
this
area
as
forested.
By
doing
so,
carbon
fluxes
are
overestimated,
in
lowland
Central
Yakutia
alone,
by
0.9
x
10
-3
Pg.
This
could
appear
as
small
but
if
other
areas,
where
thermokarst
depressions
are
present,
in
eastern
Siberia
are
added
this
value
will
become
larger.
Additionally,
these
grasslands
are
large
emitters
of
CH
4
(Takakai
et
al.,
2008),
especially
in
wet
years,
with
higher
warming
potential,
which
if
added
to
the
equation
increases
the
importance
of
taking
grasslands
into
account
for
regional
carbon
budget
models.
5.
Conclusions
Our
study
showed
that
alas
ecosystems
are
a
sink
of
carbon
but
to
a
lesser
degree
than
the
forest
(60%
less).
Despite
this
large
difference
in
the
carbon
sequestration
capacity
between
the
alas
and
the
forest,
the
forest
water
balance
is
only
13%
higher
in
the
alas
than
in
the
forest.
The
results
of
this
study
suggest
that
models
that
estimate
the
carbon
budget
in
the
Siberian
Taiga
must
take
grasslands
into
account
to
avoid
significant
overestimation.
This
study
was
conducted
during
a
wet
year
leaving
the
possibility
open
that
in
a
dry
year
carbon
uptake
and
water
vapor
fluxes
in
the
alas
ecosystem
could
be
smaller.
If
the
present
global
warming
accelerates
permafrost
degradation
in
the
forest,
the
carbon
sequestration
potential
of
this
region
could
be
drastically
reduced
as
occurred
in
the
past
when
forest
turned
into
grassland.
1976
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
Acknowledgements
This
work
was
supported
by
the
Ministry
of
Education
Sports
and
Science
of
japan
through
the
Project
RR2002.
We
wish
to
thank
Dr.
T.
Ohta,
Dr.
A.
Kotani
and
Dr.
N.
Kobayashi
for
their
constant
advise
and
support
in
the
analysis
of
the
data
and
Dr.
R.
Argunov
for
his
support
during
the
field
work.
Our
thanks
go
also
to
Dr.
A.
Cronin
for
his
advice
on
English
corrections
to
the
paper.
REFERENCES
Ameth,
A.,
Kelliher,
F.M.,
Bauer,
G.,
Hollinger,
D.Y.,
Byers,
J.N.,
Hunt,
J.E.,
McSeveny,
T.M.,
Ziegler,
W.,
Vygodskaya,
N.N.,
Milukova,
I.,
Sogachov,
A.,
Varlagin,
A.,
Schulze,
E.-D.,
1996.
Environmental
regulation
of
xylem
sap
flow
and
total
conductance
of
Larix
gmelinii
trees
in
eastern
Siberia.
Tree
Physiol.
16,247-255.
Baldocchi,
D.,
Kelliher,
F.M.,
Black,
T.A.,
Jarvis,
P.,
2000.
Climate
and
vegetation
controls
on
boreal
zone
energy
exchange.
Global
Change
Biol.
6,69-83.
Bosikov,
N.P.,
1991.
Evolution
of
alases
of
the
Central
Yakutia.
Permafrost
Institute
Yakutsk,
p.
127
(in
Russian).
Brouchkov,
A.,
Fukuda,
M.,
Fedorov,
A.,
Konstantinov,
P.,
Iwahana,
G.,
2004.
Thermokarst
as
a
short-term
permafrost
disturbance,
Central
Yakutia.
Permafrost
Periglac.
Process.
15,81-87.
Brutsaert,
W.,
1982.
Evaporation
into
the
Atmosphere
-
Theory,
History
and
Applications.
D.
Riedel,
Dordrecht,
p.
299.
Czudek,
T.,
Demek,
J.,
1970.
Thermokarst
in
Siberia
and
its
influence
on
the
development
of
lowland
relief.
Quatem.
Res.
1,103-120.
Desyatkin,
R.V.,
1993.
Syngenetic
soil
salinization
during
thermokarst
alas
formation.
Eurasian
Soil
Sci.
25,38-46.
Dolman,
A.J.,
Maximov,
T.C.,
Moors,
E.J.,
Maximov,
A.P.,
Elbers,
J.A.,
Kononov,
A.V.,
Waterloo,
M.J.,
Van
der
Molen,
M.K.,
2004.
Net
ecosystem
exchange
of
carbon
dioxide
and
water
of
far
eastern
Siberian
larch
(Larix
cajanderi)
on
permafrost.
Biogeosciences
1,133-146.
Dutta,
K.,
Schuur,
E.A.J.,
Neff,
J.C.,
Zimov,
Z.A.,
2006.
Potential
carbon
release
from
permafrost
soils
on
Northeastern
Siberia.
Global
Change
Biol.
12,2336-2351.
Fedorov,
A.N.,
Botulu,
T.A.,
Varlamov,
C.P.,
1991.
Permafrost-
landscape
Map
of
Yakutskaya
ASSR,
scale
1:2,500,000.
In:
Melnikov,
P.I.
(Ed.),
GUGK,
Moskow,
2
pp.
(in
Russian).
Hirano,
T.,
Lopez,
C.M.L.,
Kobayashi,
Y.,
Machimura,
T.,
Iwahana,
G.,
Tamura,
S.,
Kodama,
Y.,
Fedorov,
A.N.,
Fukuda,
M.,
2006.
Continuous
measurements
of
soil
respiration
at
a
larch
forest
in
eastern
Siberia.
In:
Hatano,
R.,
Guggenberger,
G.
(Eds.),
Symptom
of
Environmental
Change
in
Siberian
Permafrost
Region.
Hokkaido
University
Press,
Sapporo,
Japan,
pp.
151-157.
Hobbie,
S.E.,
Schimel,
J.P.,
Trumbore,
S.E.,
Randerson,
J.R.,
2000.
Controls
over
carbon
storage
and
turnover
in
high-latitude
soils.
Global
Change
Biol.
6,196-210.
Hollinger,
D.Y.,
Kelliher,
F.M.,
Schulze,
E.-D.,
Bauer,
G.,
Arneth,
A.,
Byers,
J.N.,
Hunt,
J.E.,
McSeveny,
T.M.,
Kobak,
K.I.,
Milukova,
I.,
Sogatchev,
A.,
Tatarinov,
F.,
Varlargin,
A.,
Zieglerf,
W.,
Vygodskaya,
N.N.,
1998.
Forest
atmosphere
carbon
dioxide
exchange
in
eastern
Siberia.
Agric.
For.
Meteorol.
90,291-306.
Hui,
D.,
Wan,
S.,
Katul,
G.,
Monson,
R.,
Luo,
Y.,
2004.
Gap-filling
missing
data
in
eddy
covariance
measurements
using
multiple
imputation
(MI)
for
annual
estimations.
Agric.
For.
Meteorol.
121,93-111.
Hunt,
J.E.,
Kelliher,
F.M.,
McSeveny,
T.M.,
Byers,
J.N.,
2002.
Evaporation
and
carbon
dioxide
exchange
between
the
atmosphere
and
a
tussock
grassland
during
a
summer
drought.
Agric.
For.
Meteorol.
111,65-82.
Iwahana,
G.,
Machimura,
T.,
Kobayashi,
Y.,
Fedorov,
A.N.,
Konstantinov,
P.Y.,
Fukuda,
M.,
2005.
Influence
of
forest
clear-cutting
on
the
thermal
and
hydrological
regime
of
the
active
layer
near
Yakutsk,
eastern
Siberia.
J.
Geophys.
Res.
110,
G02004,
doi:
10.1029/2005JG000039
Jaksic,
V.,
Kiely,
G.,
Albertson,
J.,
Oren,
R.,
Katul,
G.,
Leahy,
P.,
Byrne,
K.A.,
2006.
Net
ecosystem
exchange
of
grassland
in
contrasting
wet
and
dry
years.
Agric.
For.
Meteorol.
139,
323-334.
Jorgenson,
M.T.,
Shur,
Y.L.,
Pullman,
E.R.,
2006.
Abrupt
increase
in
permafrost
degradation
in
Arctic
Alaska.
Geophys.
Res.
Lett.
33,
L02503,
1-4,
doi:10.1029/2005GL024960.
Kachurin,
S.P.,
1962.
Thermokarst
within
the
territory
of
the
USSR.
Biuletyn
Peryglacjalny
4,49-55
(in
Russian).
Karl,
T.M.,
Knight,
R.W.,
Plummer,
N.,
1995.
Trends
in
high
frequency
climate
variability
in
the
twentieth
century.
Lett.
Nature
377,217-220.
Katamura,
F.,
Fukuda,
M.,
Bosikov,
N.P.,
Desyatkin,
R.V.,
Nakamura,
T.,
Moriizumi,
J.,
2006.
Thermokarst
formation
and
vegetation
dynamics
inferred
from
a
palynological
study
in
Central
Yakutia,
Eastern
Siberia.
Arct.
Antarct.
Alp.
Res.
38
(4),
561-570.
Kelliher,
F.M.,
Leuning,
R.,
Schulze,
E.-D.,
1993.
Evaporation
and
canopy
characteristics
of
coniferous
forests
and
grasslands.
Oecologia
95,153-163.
Kelliher,
F.M.,
Hollinger,
D.Y.,
Schulze,
E.-D.,
Vygodskaya,
N.N.,
Byers,
J.N.,
Hunt,
J.E.,
McSeveny,
T.M.,
Milukove,
I.,
Sogatchev,
A.,
Varlargin,
A.,
Ziegler,
W.,
Ameth,
A.,
Bauer,
G.,
1997.
Evaporation
from
an
eastern
Siberian
larch
forest.
Agric.
For.
Meteorol.
85,135-147.
Kimura,
R.,
Fan,
J.,
Zhang,
X.,
Takayama,
N.,
Kamichika,
M.,
Matsuoka,
N.,
2006.
Evapotranspiration
over
the
grassland
Field
in
the
Liudaogou
Basin
of
the
Loess
Plateau,
China.
Acta
Oecologica
29,45-53.
Korman,
R.,
Meixner,
F.X.,
2001.
An
analytical
footprint
model
for
non-neutral
stratification.
Bound.
Layer
Meteorol.
99,
207-224.
Kuwada,
T.,
Kokake,
T.,
Takeuchi,
S.,
Maximov,
T.C.,
Yoshikawa,
K.,
2002.
Relationships
among
water
dynamics,
soil
moisture
and
vapor
pressure
deficit
in
a
Larix
gmelinii
stand,
eastern
boreal
Siberia.
J.
Jpn.
Forest
Soc.
84
(4),
246-
254
(in
Japanese).
Lopez,
C.M.L.,
Brouchkov,
A.,
Nakayama,
H.,
Takakai,
F.,
Fedorov,
A.N.,
Fukuda,
M.,
2007a.
Epigenetic
salt
accumulation
and
water
movement
in
the
active
layer
of
Central
Yakutia
in
Eastern
Siberia.
Hydrol.
Process.
21,
103-109.
Lopez,
C.M.L.,
Saito,
H.,
Kobayashi,
Y.,
Shirota,
T.,
Iwahana,
G.,
Maximov,
T.C.,
Fukuda,
M.,
2007b.
Inter-annual
environmental-soil
thawing
rate
variation
and
its
control
on
transpiration
from
Larix
cajanderi,
Central
Yakutia
Eastern
Siberia.
J.
Hydrol.
338,251-260.
Machimura,
T.,
Kobayashi,
Y.,
Iwahana,
G.,
Hirano,
T.,
Lopez,
C.M.L.,
Fukuda,
M.,
Fedorov,
A.N.,
2005.
Change
of
carbon
dioxide
budget
during
three
years
after
deforestation
in
Eastern
Siberian
Larch
Forest.
J.
Agric.
Meteorol.
60,
653-656.
Muller,
M.,
1982.
Selected
climatic
data
for
a
global
set
of
standard
stations
for
vegetation
science.
Dr.
W.
Junk
Publishers,
The
Hague,
p.
306.
Ohta,
T.,
Hiyama,
T.,
Tanaka,
H.,
Kuwada,
T.,
Maximov,
T.C.,
Ohata,
T.,
Fukushima,
Y.,
2001.
Seasonal
variation
in
the
energy
and
water
exchanges
above
and
below
a
larch
forest
in
eastern
Siberia.
Hydrol.
Process.
15,
1459-1476.
AGRICULTURAL
AND
FOREST
METEOROLOGY
148
(2008) 1968-1977
1977
Osterkamp,
T.E.,
Romanovsky,
V.E.,
1999.
Evidence
for
thawing
of
discontinuous
Permafrost
in
Alaska.
Permafrost
Periglac.
Process.
10,17-37.
Payette,
S.,
Delwaide,
A.,
2000.
Recent
permafrost
dynamics
in
a
subarctic
floodplain
associated
with
Changing
Water
Levels,
Quebec,
Canada.
Arct.
Antarct.
Alp.
Res.
32,316-323.
Romanovsky,
V.E.,
Osterkamp,
T.E.,
Duxbury,
N.S.,
1997a.
An
evaluation
of
three
numerical
models
used
in
simulations
of
the
active
layer
and
permafrost
temperature
regimes.
Cold
Regions
Sci.
Technol.
26,195-203.
Saito,
H.,
Shirota,
T.,
Lopez,
C.M.L.,
Iwahana,
G.,
Mximov,
T.C.,
Takahashi,
K.,
2005.
Changing
precipitation
regimes
and
leaf
area
in
a
canopy
of
mature
larch
forest
in
eastern
Siberia.
In:
Ohata,
T.,
Ohta,
T.,
Kodama,
Y.
(Eds.),
Proceedings
for
2nd
Symposium
on
Water
Cycle
in
Northern
Eurasia.
p.
53.
Shender,
N.I.,
Romanovsky,
B.E.,
Tetelbaum,
A.C.,
1999.
Prognosis
of
natural
historical
climate
fluctuations
of
Yakutsk
and
Fairbanks.
Nauka
y
Obrazobanie
2,24-29
(in
Russian).
Shvidenko,
A.,
Nilsson,
S.,
1994.
What
do
we
know
about
the
Siberian
forest?
Ambio
23,389-404.
Smith,
M.W.,
1975.
Microclimatic
influences
on
Ground
Temperatures
and
Permafrost
Distribution,
Mackenzie
Delta
Northwest
Territories.
Can.
J.
Earth
Sci.
12,1421-1438.
Soussana,
J.F.,
Allard,
V.,
Pilegaard,
K.,
Ambus,
P.,
Amman,
C.,
Campbell,
C.,
Ceschia,
E.,
Clifton-Brown,
J.,
Czobel,
S.,
Domingues,
R.,
Flechard,
C.,
Fuhrer,
J.,
Hensen,
A.,
Horvath,
L.,
Jones,
M.,
Kasper,
G.,
Martin,
C.,
Nagy,
Z.,
Neftel,
A.,
Raschi,
A.,
Baronti,
S.,
Rees,
R.M.,
Skiba,
U.,
Stefani,
P.,
Manca,
G.,
Sutton,
M.,
Tuba,
Z.,
Valentini,
R.,
2007.
Full
accounting
of
the
greenhouse
gas
(CO
2
,
N
2
0
CH
4
)
budget
of
nine
European
grassland
sites.
Agric.
Ecosyst.
Environ.
121,
121-134.
Sugimoto,
A.,
Naito,
D.,
Yanagisawa,
N.,
Ichiyanagi,
K.,
Kurita,
N.,
Kubota,
J.,
Kotake,
T.,
Ohata,
T.,
Maximov,
T.C.,
Fedorov,
A.N.,
2003.
Characteristics
of
soil
moisture
in
permafrost
observed
in
East
Siberian
taiga
with
stable
isotopes
of
water.
Hydrol.
Process.
17,1073-1092.
Takakai,
F.,
Desyatkin,
A.R.,
Lopez,
C.M.L.,
Hatano,
R.,
Fedorov,
A.N.,
Desyatkin,
R.V.,
2006.
Greenhouse
gas
emissions
from
a
Siberian
alas
ecosystem
near
Yakutsk
Russia.
In:
Hatano,
R.,
Guggenberger,
G.
(Eds.),
Symptom
of
Environmental
Change
in
Siberian
Permafrost
Region.
Hokkaido
University
Press,
Sapporo,
Japan,
pp.
11-25.
Takakai,
F.,
Desyatkin,
A.R.,
Lopez,
C.M.L.,
Fedorov,
A.N.,
Desyatkin,
R.V.,
Hatano,
R.,
2008.
CH
4
and
NO
2
emissions
from
a
forest-alas
ecosystem
in
the
permafrost
taiga
forest
region,
eastern
Siberia,
Russia.
J.
Geophys.
Res.
113,
G02002,
doi:10.1029/2007JG000521,
2008.
Webb,
E.K.,
Pearman,
G.I.,
Leuning,
R.,
1980.
Correction
of
flux
measurements
for
density
effects
due
to
heat
and
water
vapour
transfer.
Q
J.
R.
Meteorol.
Soc.
106,85-100.
Wilson,
K.,
Goldstein,
A.,
Falge,
E.,
Aubinet,
M.,
Baldocchi,
D.,
Berbigier,
P.,
Bemhofer,
C.,
Ceulemans,
R.,
Dolman,
H.,
Field,
C.,
Grelle,
A.,
Ibrom,
A.,
Law,
B.E.,
Kowalski,
A.,
Meyers,
T.,
Moncrieff,
J.,
Monson,
R.,
Oechel,
W.,
Tenhunen,
J.,
Valentini,
R.,
Verma,
S.,
2002.
Energy
balance
closure
at
FLUXNET
sites.
Agric.
For.
Meteorol.
113,223-243.