taking the pulse of mountain forests · basal area gain. basal area loss. difference. year 1960...
TRANSCRIPT
U.S. Department of the InteriorU.S. Geological Survey
TAKING THE PULSE OFMOUNTAIN FORESTS:
THE CORDILLERA FORESTDYNAMICS NETWORK
(CORFOR)
CORFOR is a collaboration of scientists with permanent foreststudy plots arrayed along the American Cordillera
Nate Stephenson U.S. Geological Survey, USAAlvaro Duque Univ. Nacional de Colombia, MedellínEsteban Alvarez Interconexión Eléctrica S.A., ColombiaJulieta Carilla Univ. Nacional de Tucumán, ArgentinaLori Daniels Univ. of B.C., Vancouver, CanadaRicardo Grau Univ. Nacional de Tucumán, ArgentinaGreg Greenwood Mountain Res. Initiative, SwitzerlandMark Harmon Oregon State Univ., Corvallis, USASergio Orrego Univ. Nacional de Colombia, MedellínPhil van Mantgem U.S. Geological Survey, USATom Veblen University of Colorado, Boulder, USA
Part 1: Information needs: mountain forests in achanging climate
Part 2: Permanent forest plots uniquely meet some ofour information needs
Part 3: Need for, and birth of, the Cordillera Forest Dynamics Network (CORFOR)
Part 4: CORFOR progress report
Road map:
Mountain forests provide:• Wood and other products• Hydrologic regulation• Recreational and spiritual
opportunities• Critical wildlife habitat and
biodiversity conservation
Mountain forests provide:• Wood and other products• Hydrologic regulation• Recreational and spiritual
opportunities• Critical wildlife habitat and
biodiversity conservation
With rapid climatic change,we potentially face:• Changes (included losses) in
some of these goods andservices
• Large feedbacks to climaticchange
If we hope to adapt to and mitigatethe effects of climatic changes on forests,
we have at least two critical needs:
(1) Ability to detect ongoing changes.
(2) Ability to forecast future changes.
Hicke et al. 2002, Global Biogeochem. Cycles
2. PERMANENT FOREST PLOTS UNIQUELYMEET SOME OF OUR INFORMATION NEEDS
Change detection:
Advantages of remote sensing:• Wall-to-wall coverage across a range of spatial scales!
Need for plot-based data:• Ground-truthing of remotely-sensed data• Detecting things not detected by remote sensing
- Changes in reproduction- Subtle changes in mortality rates- Ongoing compositional shifts
The two approaches are complementary!
Change detection:
Tropical (Amazonian) forest DYNAMICS are changing(e.g., recruitment, growth, and mortality rates are increasing)
Phillips et al., Phil. Trans. B, 2004
Recruitment
Mortality
Phillips et al., Nature, 2002
Credit: Yadvinder Malhi
Tropical (Amazonian) forest COMPOSITION is changing(e.g., lianas [woody vines] are increasing)
Lewis et al., Phil. Trans. B, 2004
Tropical (Amazonian) forest STRUCTURE AND FUNCTIONare changing (e.g., aboveground biomass, hence C storage,
may be increasing)
Basal area gainBasal area loss
Difference
Year1960 1970 1980 1990 2000
Mor
talit
y ra
te (%
yr-1
)
0.0
0.5
1.0
1.5
2.0P<0.0001, n=76
Temperate (western USA) forest DYNAMICS are changing(background mortality rates are increasing)
van Mantgem et al. 2008
Year1990 2000 2010 2020 2030 2040 2050
PIL
A p
opul
atio
n si
ze
0
100
200
300
400
500observedprojection using 83-88 dataprojection using 88-94 dataprojection using 94-2000 data
van Mantgem et al. Ecol. Appl. 2004
Temperate (western USA) forest COMPOSITION is changing(e.g., 5-needled pines are declining)
Temperate forest STRUCTURE AND FUNCTIONare almost certainly changing, but …
Trust us,George … we’re
working on it!
Temperate forest STRUCTURE AND FUNCTIONare almost certainly changing, but …
Change detection:Permanent forest plots provide our
“Keeling curve” for forests
Most currentmodels are driven
by sets of untestedassumptions,
many of which aredecades old.
Forecasting:
Growth rate
Pro
babi
lity
of d
eath
The well-known inverse relationship between tree growth rate and probability of death is at the heart of “gap” models.
Growth rate
Pro
babi
lity
of d
eath
A key model assumption: The only way climatic changes affect probability of death (in the absence of catastrophic
disturbance) is by altering tree growth rate.
How do we test assumptions like this when thegreat spatial and temporal scales of forest dynamics
preclude experimentation?
Replicatedmountain
forestsin bell jars
We take advantage of natural experimentsin space and time … permanent forest plots
arrayed along environmental gradients
Credit: A. Caprio
Elevation (m)1500 2000 2500 3000 3500
Fore
st t
urno
ver
(% y
r-1)
0
1
2
3y = 2.76 - 0.00066 xr 2 = 0.49, P < 0.001
In the forests of California’s Sierra Nevada,forest turnover rate (mortality and recruitment)
declines significantly with elevation.
Stephenson & van Mantgem, Ecol. Lett., 2005
0
1
2
3
4
550 46 30 27 27
Rich
er s
oils
Poor
er s
oils
Angi
ospe
rm
Gym
nosp
erm
Mixe
d
Tropical Temperate(Amazonia) (global)
Fore
st t
urno
ver
(% y
r-1)
Stephenson & van Mantgem 2005
Fore
st t
urno
ver
(% y
r-1)
0
1
2
3
4
5158 84
Trop
ical
Tem
pera
te
• Globally, forests of productive environments have higher turnover rates …
• ... at least partly because environments that favor tree growth alsofavor the organisms that kill trees.
Natural experiments in space suggestthe need for a fundamental change in the nature
of the gap model mortality function.
Mor
talit
y ra
te
Site productivityIndividual treegrowth rate
Low
High Low
High
Summary: some roles of permanent forest plots
(1) Change detection
(2) Provide a mechanisticfoundation forforecasting
Year1960 1970 1980 1990 2000
Mor
talit
y ra
te (%
yr-1
)
0.0
0.5
1.0
1.5
2.0P<0.0001, n=76
Mor
talit
y ra
te
Site productivityIndividual treegrowth rate
Low
High Low
High
3. NEED FOR, AND BIRTH OF, THE CORDILLERAFOREST DYNAMICS NETWORK (CORFOR)
The RAINFOR plot networkhttp://www.geog.leeds.ac.uk/projects/rainfor/
Only two other broad, international forest dynamics networksexist, and both focus on lowland tropical forests
The Center for Tropical Forest Science (CTFS) plot networkhttp://www.ctfs.si.edu/
CORFOR’s vision: a network taking advantage of thebroadest environmental gradients possible
Latitudes: Subarctic – temperate – tropical (and back).Temperature (elevation): Sea level to >3500 m.Precipitation: West-to-east gradients.Soils and topography: Local variation.
CORFOR’s vision: a network taking advantage of thebroadest environmental gradients possible
Based on this vision, CORFOR was born at the CONCORDclimate change conference in Mendoza, Argentina,
4 - 6 April 2006
4. CORFOR PROGRESS REPORT
CORFOR workshop,13 June 2008, thanks to:
http://www.corfor.com/http://mri.scnatweb.ch/content/view/88/67/
- Mountain ResearchInitiative (MRI),Switzerland
- USGS Western MountainInitiative (WMI)
- MTNCLIM 2008organizers
Among other things, tomorrow’s workshop will:
- Establish a set of minimum standards for forestdynamics plots in the network.
- Create a road map for the future: identifyresearch needs and opportunities of broadsignificance that CORFOR can address.
- As a “proof of concept,” work on a papercomparing forest dynamics along elevationalgradients in the southern temperate zone,tropics, and northern temperate zone.
Northern temperate zone:California,Oregon,
and Washingtonplot networks
California
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.54, P < 0.0001
Mortality rate vs. elevation, northern temperate zone
California
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.54, P < 0.0001
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.08, P = 0.08
Oregon & Washington
Mortality rate vs. elevation, northern temperate zone
.....
.
...
.
.
.
.
5-ha
25-ha
2-ha
1-ha
Tropics:Colombian
plot network
Colombia
Mortality rate vs. elevation, tropics
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.31, P = 0.07
Colombia Venezuela
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.09, P = 0.24
Mortality rate vs. elevation, tropics
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.31, P = 0.07
Southern temperate zone:Tucumán, Argentina plot network
Argentina,mature forest
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.08, P = 0.42
Mortality rate vs. elevation, southern temperate zone
Argentina,mature forest
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.08, P = 0.42
Elevation (m)0 1000 2000 3000
Mea
n m
orta
lity
rate
(% y
r-1)
0
1
2
3
4R2 = 0.20, P = 0.14
Argentina,secondary forest
Mortality rate vs. elevation, southern temperate zone
CORFOR Challenges:
• Time: All of us have many other pressing obligations.
• Funds: It is very difficult to capture funds for long-term forest research, much less to hire someone tocoordinate CORFOR activities.
• Communication: We are literally continents apart. Weaccomplished more in a few hours together yesterdaythan we had during weeks of email exchanges.
CORFOR Challenges:
• Time: All of us have many other pressing obligations.
• Funds: It is very difficult to capture funds for long-term forest research, much less to hire someone tocoordinate CORFOR activities.
• Communication: We are literally continents apart. Weaccomplished more in a few hours together yesterdaythan we had during weeks of email exchanges.
Happily, we have a dedicated, motivated, andexcited core group of scientists, ready to acceptthese challenges.
Thank you for your attention!