Transcript PowerPoint Presentation - Template 02

Millennial-scale Dynamics of Continental
Peatlands in Western Canada:
Pattern, Controls and Climate Connection
Zicheng Yu
Lehigh University
Bethlehem, Pennsylvania
QUEST Workshop on CH4 & Wetlands
14-16 June 2004, Bristol, UK
Acknowledgements


Dale Vitt, Kel Wieder, Merritt Turetsky, Dave
Beilman, Ilka Bauer, Mike Apps, Celina
Campbell, and Ian Campbell for sharing
slides, data and ideas.
Climate Change Action Fund (Canada) and
National Science Foundation (US) for funding.
Outline of Talk

Overview of continental peatlands in
western Canada

Accumulation pattern & trajectories

Possible climate & global C cycle
connections

Conclusions
Peatland Types in Western Canada
Bogs (treed)
Permafrost
peatlands
Open fens
Treed fens
Peatland Distribution
% Cover
Total peatland area = 365,160 km2 (21% landbase)
63% fens
28% permafrost bogs
9% non-permafrost bogs
Vitt et al. (2000)
Peatland Carbon Storage
14
C Storage (Pg)
12
10
8
6
4
2
Parkland
Mid-boreal
High boreal
Montane
Subarctic
Arctic
Total =
48 Pg
0
Vitt et al. (2000)
Fens are more important C pool and
have larger area than bogs in continental
Canadian peatlands, as well as bigger
CH4 emitters,
but we know much less about these
ecosystems than bogs in general
Outline of Talk

Overview of continental peatlands in
western Canada

Accumulation pattern & trajectories

Possible climate & global C cycle
connections

Conclusions
Why accumulation pattern matters?
Exponential
Cumulative Mass (g.cm-2)
120
(Convex)
100
80
Linear
60
40
(Concave)
20
Logarithmic
0
0
4
Time (ka)
8
12
Because:
Observed pattern

Infer & understand the
processes

Projecting future
dynamics/trajectories
Cumulative peat mass (g/cm2)
Concave Pattern from Oceanic Bogs
0
Draved Mose, Denmark
(data from Aaby & Tauber, 1975)
5
(assuming constant
PAR and decay)
10
15
20
25
0
1000
2000
3000
4000
5000
Age (calendar year BP)
“apparent” C accumulation rate
6000
7000
Study Sites
5 sites with hiresolution peat
core analysis
Basal dates from ~80
paludified peatlands
Loss-on-Ignition from Upper Pinto Fen
1-cm LOI
n=20 dates
also,
2-cm macro
2-cm isotopes
Yu et al. 2003
Peat Depth-Age Curve: Convex at UPF
Opposite to
welldocumented
“concave”
pattern
Yu et al. 2003
What Does This Indicate?
0
Cumulative Peat Mass (g/cm2)
UPF: Convex Pattern
10
20
30
0
Causes?
1000
2000
3000
4000
5000
6000
Age (cal yr BP)
• decreasing peat-addition rates from acrotelm, and/or
• increasing catotelm decomposition rate
A Simple Extended Model
Followed the suggestion by Clymo (2000; Quebec Meeting),
dM
 p * e b*t   * M ,
dt
where M = cumulative peat mass;
p = eventual PAR;
 = catotelm decomposition rate; and
b = PAR coefficient.
This equation has an analytical solution,
p
M (
) * (e b*t  e  *t ) .
 b
Yu et al. 2003
Peat Mass (g/cm2)
Sensitivity to Changes in Decay & PAR
0
0
10
10
-50% PAR
+50% Decay
20
20
30
30
40
40
-50% Decay
50
50
+50% PAR
60
60
0
1000
2000
3000
4000
Age (cal BP)
5000
6000
0
1000
2000
3000
4000
5000
6000
Age (cal BP)
Yu et al. 2003
Change in PAR over Time
1.0
PAR modifier = exp[-b*t]
Change in PAR
PAR0.8 decrease from initial 192
to eventual 26 g/m2/yr could
explain
the observed pattern
0.6
-1
191.8 g m-2 yr-1
200
PAR Modifier
Peat-Addition Rate (g m -2yr-1)
250
150
100
0.000185 yr
(-50% b)
0.4
0.2
50
26.0 g
m-2
yr-1
0.0
0
0
2000
4000
6000
Age (cal BP)
0.000555 yr-1
(+50% b)
0
2000
b=0.00037 yr-1
4000
6000
Time (years)
Yu et al. 2003
Summary I



The model suggests that unidirectional decrease
of PAR from 192 to 26 g m-2 yr-1 over that 5400-yr
period at UPF could result in the observed convex
pattern.
Autogenic drying trend resulted from fen height
growth gradually isolates peat surface from water
and nutrient sources, causing decreased production,
especially for water-demanding rich fen species - esp.
in moisture-limiting continental regions.
This analysis indicates that continental peatlands with
convex pattern may reach their growth limit
sooner than previous model predicts.
Convex Pattern @ Other Sites I
(Kubiw et al. 1989)
Convex Pattern @ Other Sites II
Western Canada:
Slave Lake Bog (Kurry & Vitt 1996)
Southwestern Finland:
Pesansuo raised bog (Ikonen, 1993)
Western Siberia:
Salym-Yugan Mire (Turunen et al. 2001)
Convex Pattern from Regional Sites
(Yu & Vitt, in prep)
Outline of Talk

Overview of continental peatlands in
western Canada

Accumulation pattern & trajectories

Possible climate & global C cycle
connections

Conclusions
Climate Proxy from UPF
(Yu et al. 2003)
UPF
(Yu et al. 2003)
W. Canada
Global Climate & C Cycle Connections?
Bond et al.
2001
Yu et al. 2003
Indermuhle et al.
1999
Chappellaz et al.
1997
Brook et al. 2000
Summary II



Peat accumulation in western Canada shows sensitive
response to Holocene climate variability at
millennial time scale.
Peatland carbon dynamics may connect to change in
atmospheric CO2 concentrations (Peatlands in
western Canada contain ~50 Pg C, which is
equivalent to ~25 ppm CO2 if all remained in the
atmosphere).
Are there similar pattern in other peatlands of
northern latitudes?
Pervasive Climate Controls of Peatland Dynamics
A thawed bog
shows similar
millennial-scale
variations
Patuanak Bog
(internal lawn)
Connection of Siberian Peatland Initiations
and Atmospheric CH4
N = ~200
Smith et al. 2004
Bill Ruddiman’s hypothesis:
 CO2 increase since 8 ka:
caused by deforestation;
 CH4 increase since 5 ka:
caused by rice cultivation
Allogenic and Autogenic Controls of Peatland
Dynamics: a conceptual model
Autogenic drying
Climate wetting
Climate fluctuations
Yu et al. 2003
Conclusions
• The different accumulation pattern observed in
continental peatlands suggests these peatlands follow
different trajectories historically and may respond to
climate change differently (compared to well-studied bogs);
• Continental peatlands appear to show sensitive
responses to subtle millennial-scale moisture changes
during the Holocene;
• Fens seem to be more variable in C accumulation and
more sensitive (less self-regulating) to climate variations
than bogs;
• Northern peatlands might have had detectable impacts
on atmospheric CO2 and CH4 concentrations during the
Holocene.
Suggestions
• Develop scaling-up models to take advantage of detailed
inventory results from western Canada or other regions for
regional CH4 emission estimates by peatland types (as a
validating tool for global model?);
• Confirm the extent of past climate – peatland – global C cycle
connections, particularly using multiple proxies from paired lakepeatland approach (lakes for independent climate
reconstructions);
• Understand implications of permafrost (intact, thawing, and
thawed) peatlands (and fen-bog transition) for CH4
emission/budget – permafrost is one of the biggest surprises to
come in peatland C dynamics;
• Integrate/reconcile down-core paleo data with present-day
instrumental C flux measurements.