Title: Old
1Examining Changes in Forest Ecosystem Carbon
Exchange with Stand Age in the Upper Midwest US
Old
Not So Old
Ankur Desai, Ken Davis, Weiguo Wang Department
of Meteorology Pennsylvania State
University Paul Bolstad, Eileen Carey, Bruce
Cook Jon Martin, Leslie Kreller, Deborah
Hudleston Department of Forest
Resources University of Minnesota
2Acknowledgements
- Department of Energy
- Office of Biological and Environmental Research,
Terrestrial Carbon Processes - National Institute for Global Environmental
Change (NIGEC) Midwestern Region - Ameriflux / Fluxnet
- U.S. Forest Service
- Chequamegon-Nicolet National Forest, Park Falls,
WI - Ottawa National Forest, Watersmeet, MI
- North Central Research Station, Rhinelander, WI
3Motivation
- While interannual variability in forest carbon
exchange appears to be primarily controlled by
climate, long-term forest carbon exchange is
strongly constrained by stand age since
initiation - Scaling carbon fluxes across space and time (past
or future) requires understanding of the
relationship between forest carbon exchange and
site age - Eddy flux measurements in old-growth forests are
few but increasing. None are in the upper
Midwest - Old growth forests are expanding in the upper
Midwest - Old growth forests are not carbon neutral as was
typically theorized, but are probably small
carbon sinks
4ChEAS Region
- Chequamegon-Ecosystem Atmosphere Study allows
for intensive study of carbon exchange - Most forests in region arerecovering from
logging in early 20th century - Pre-European settlement vegetation is found in
the 8,500 Ha Sylvania Wilderness and Recreation
area (Upper Peninsula, Michigan) - Eddy flux towers have been installed in an old
growth stand adjacent to Sylvania and a nearby
mature hardwoods forest (Willow Creek)
5Hypotheses
- Ecosystem respiration is greater at an old-growth
forest compared stand to a mature forest stand - Primarily due to increased coarse woody debris
- Gross ecosystem production is slightly lower at
the old-growth site compared to the mature site - Possibly due to decreased light use efficiency,
stomatal conductance, water use efficiency and/or
hydraulic conductivity - The old-growth site is more sensitive to drought
stress
6Site Descriptions
- Sylvania old-growth
- Established late 2001
- Hemlock, sugar maple, basswood, yellow birch,
LAI 3.6 - 0-350 year old trees, very low disturbance
frequency (5.4 / decade), long-term
compositional equilibrium - Lots of coarse woody debris (42 Mg/Ha)
- 3-30 Ha patchy regions of s. maple or hemlock
dominance - Willow Creek mature hardwoods
- Established mid-1998
- Sugar maple, basswood, green ash, LAI 5.3
- Approx. 70 yrs old (initiated after logging)
- Very little coarse woody debris
7Site Descriptions
- Willow Creek pre-European settlement vegetation
is very similar to Sylvania
8Site Descriptions
9Climate
- Average (1971-2000, Watersmeet, MI)
- Annual 3.9o C, -12.7o C in Jan, 18.5o C in Jul
- 771 mm of precipitation (57 during May-Sep)
- Willow Creek is slightly colder and drier (esp.
in winter) - 2002
- Warmer 5.0o C, -7.1o C in Jan, 20.7o C in Jul
- Wetter 866 mm, but very dry in Nov and Dec
- 2003
- Colder up through Oct -14.3o C in Jan, 18.0o C
in Jul - Very dry winter and spring (50 of average
precip.) - August-October also dry (59 of average precip.)
- Soil moisture lower than 2002 in Jan-Mar and
Jul-Sep
10Climate
11Methods
- Eddy covariance measurements and high-precision
CO2 mixing ratio profiles measured from 30 m (W
Creek) and 37 m (Sylvania) towers, 10 hz, similar
instrumentation (closed-path IRGAs) and flux
calculation methods - Standard micrometeorological and soil
measurements and profiles (temperature, water,
radiation, etc) - Ground-based biometric and physiological
measurements in vicinity of towers and across
ChEAS region, including soil efflux
12Methods
- Nighttime u cutoff filter
- 0.3 m s-1 at Sylvania
- 0.175 m s-1 at Willow Creek
- Directional heterogeneity NEE data screening
- non-SW winds on summer nights at Sylvania
(dilution by lakes?) - SE-winds (typically with temperature inversions)
at Willow Creek (flux pooling in low-lying
wetland?)
13Methods
- Ecosystem respiration (ER) modeled with nighttime
NEE and 5-cm soil temperature - Using an Arrhenius style equation
- 28-day moving window fit
- GEP computed as ER NEE
- Light response curves modeled with daytime GEP
and above canopy incoming total PAR - Rectangular hyperbola curve
- Similar moving window fit as for ER
- Data gaps filled using ER and GEP models
- 2003 results are preliminary (computed last week!)
14Average Daily NEE (Jun-Aug)
- Mean nighttime summer NEE at Willow Creek was
smaller than Sylvania nighttime NEE decreased at
Sylvania in 2003 vs. 2002, but not at Willow
Creek - Mean daytime summer NEE at Willow Creek was
larger than Sylvania daytime NEE was similar in
02 and 03
15Net Ecosystem Exchange
-
- Total summer (Jun-Aug) NEE at Willow Creek was
larger than Sylvania 244 larger in 2002, 181
in 2003 - Summer Sylvania NEE increased 60 in 2003 vs.
2002, while Willow Creek NEE increased 30
gC m-2 day-1
gC m-2 day-1
16Cumulative NEE
- Annual total NEE in 2002 at Willow Creek (-447 gC
m-2 yr-1) was much larger than Sylvania (-72 gC
m-2 yr-1) - 2003 is also on target for a large difference up
through September, NEE Jan-Oct 2003 at Willow
Creek (-701 gC m-2 yr-1) was 1.75x larger than
Sylvania (-254 gC m-2 yr-1). Total uptake
appears to be larger in 2003 vs. 2002
17Respiration Response
-
- Sylvania had a greater temperature dependency
(dR/dT) to respiration than Willow Creek - Respiration for a given temperature decreased
slightly for Sylvania, but increased for Willow
Creek in 2003 - Summer soil temperature was 1.2 C cooler in 2003
18Total Ecosystem Respiration
-
- Summer respiration at Willow Creek was smaller
than Sylvania 41 smaller in 2002, 24 in 2003 - Summer Sylvania respiration decreased 26 in 2003
vs. 2002, whereas Willow Creek declined only 3
Total Sylvania (965 gC m-2 yr-1) Willow Creek
(667 gC m-2 yr-1)
gC m-2 day-1
gC m-2 day-1
19Soil Chamber Efflux
- Sylvania
- Total 2002 soil respiration was 908 gC m-2 yr-1
94 of total ecosystem respiration (965 gC m-2
yr-1) - Mean summer soil respiration was 3.9 mmol m-2 s-1
57 of mean summer eddy flux respiration (6.8
mmol m-2 s-1) - Willow Creek
- Total 2002 soil respiration was 1147 gC m-2 yr-1
172 of total ecosystem respiration (667 gC m-2
yr-1) - Mean summer soil respiration was 7.0 mmol m-2 s-1
175 of mean summer eddy flux respiration (4.0
mmol m-2 s-1)
20Daytime PAR Response
- GEP to PAR response was slightly smaller at
Sylvania than Willow Creek in 2002 bigger
difference in 2003 - Willow Creek response curves were relatively
similar in 2002 and 2003 Sylvania was smaller in
2003 vs. 2002
21Total Ecosystem GEP
-
- Summer GEP at Willow Creek was slightly larger
than Sylvania 15 larger in 2002 46 in 2003 - Summer Sylvania GEP declined 7 in 2003 vs 2002,
whereas Willow Creek GEP increased 17
Total Sylvania (1036 gC m-2 yr-1) Willow Creek
(1136 gC m-2 yr-1)
gC m-2 day-1
gC m-2 day-1
22GEPER
- Mean daily GEPER ratio for summer (Jun-Aug)
2002 2003
Willow Creek 2.5 /- 0.5 2.9 /- 0.5
Sylvania 1.3 /- 0.3 1.8 /- 0.9
23WUE
- Ratio of mean monthly GEP (gC m-2 day-1) LE (kg
H2O m-2 day-1) - Significant change from 2002 to 2003 at Willow
Creek - LE decreased significantly in summer at Willow
Creek, but not at Sylvania from 2002 to 2003
June July August
Willow Creek
2002 4.1 3.2 3.5
2003 6.1 6.6 6.0
Sylvania
2002 3.4 3.6 3.5
2003 2.7 3.6 3.9
24Summary
- Ecosystem respiration at Sylvania appears to be
more sensitive to climate than Willow Creek.
Both Willow Creek and Sylvania had significant
GEP interannual variability, though more at
Willow Creek - The two sites had opposite responses in summer
GEP when comparing 2002 to 2003 - Total annual NEE at both sites was larger in a
cool, dry year (2003) compared to a warm, wet
year (2002) - This NEE increase was driven at Sylvania mainly
by ecosystem respiration decline, but at Willow
Creek, the increase was due to an increase in GEP
25Summary
- Chamber soil efflux suggests opposite respiration
differences between the two sites than eddy flux - CWD respiration calculations underway
- GEPER increased slightly at both sites in 2003
compared to 2002 - WUE increased at Willow Creek in 2003 vs. 2002
- Sapflux data can be used to separate evaporation
and transpiration to find source of WUE change - Had Willow Creek never been logged or managed, it
probably would have much larger ecosystem
respiration (most likely due to greater CWD), and
slightly or moderately smaller gross ecosystem
production
26Future Work
- Analysis of biometric and physiological
measurements to calculate NEP and individual
carbon pool fluxes - BIOME-BGC modeling of carbon exchange
- Integration of remote sensing / FIA data to
extrapolate fluxes over the region - More information at http//cheas.psu.edu