Transcript Biophysical controls - Landscape Ecology, Modeling, Mapping and

Multiscale Climatic, Topographic, and Biotic Controls of Tree Invasion in a
Sub-Alpine Parkland Landscape, Jefferson Park,
Oregon Cascades, USA
Harold S.J. Zald, Oregon State University
MTNCLIM 2010 | 06.09.2010
HJA | Blue River | OR
Why Study Tree Invasion in Mountain Ecotones?
• Globally, treeline positions related to thermal conditions
• Treeline movement a highly variable response to climate
across multiple climate regions, species, & land use histories
Harsch et al. (2009) Ecology Letters
• Treeline movement and meadow encroachment may influence:
ecosystem productivity, carbon balance, energy budget, hydrological
processes, species distributions, and biodiversity
Multiscale Drivers of Tree Invasion
• Tree invasion
fundamentally driven by
regeneration processes
Multiple
Gradient
Subalpine
Parkland
Edaphic
“TripleGradient
Treeline”
Gentle
Elevation
Treeline
• Not just climate!
• Biophysical controls:
topography, soils,
disturbance, seed
sources, facilitation,
competition, etc.
• Biophysical factors can
control spatial patterns &
sensitivity to climate
Mount
Hood,
Denali
Banff National
National
Park,
Park,Oregon
Canada
Alaska
Spatial Autocorrelation of Biophysical Controls
• Traditionally, observations of treeline movement &
meadow invasion along transects with limited environmental
gradients
• Biophysical variables
spatially autocorrelated
• Difficult to untangle
drivers
Modified from
Brooke et al. 1970
Applying New Technologies to Old Questions
• Light Detection and Ranging
(LiDAR)
• Landscape characterization of
fine-scale vegetation structure &
topography
• LiDAR can be used to sample
across multiple biophysical
gradients at scales compatible
regeneration processes
0
50
100 m
Research Questions
• How have climatic and
biophysical factors
controlled recent
rates & spatial patterns
of tree invasion?
Background:
• PNW tree invasion driven by
snow depth and persistence,
believed to determine growing
season length
(Franklin et al. 1971, Woodward et al.
1995, Rochefort & Peterson 1996)
Study Area: Jefferson Park, OR
• Mount Jefferson Wilderness
Willamette NF
• Elevation: 1755-1840 m
• ~130 ha
• Tree islands of mountain
hemlock & Pacific silver fir
• Geomorphology: glacial &
neoglacial debris flows
• No known fires
• Unknown grazing history
Jefferson
Park
HJA
LiDAR Derived Biophysical Variables
Bio
Overstory canopy
Influences snow depth & persistence,
seed sources
Physical
Topographic position, elevation, radiation
Influence snow depth & persistence
Landform (glacial v. debris flows)
Disturbance, but also influences other biophysical
variables
LiDAR Driven Sampling
Sample along biophysical
gradients believed to
influence snow depth
& persistence
100 x 100 m cluster
Topography (5 Classes)
Distance to overstory canopy
(5 classes)
Combine grids
(5 x 5 = 25 classes)
100 x 100 m moving window
(20 clusters)
Stratified random sample
(25 points per cluster)
LiDAR Driven Sampling Continued
• Mapping of overstory
canopy by species
(potential seed sources)
• Spatial autocorrelation
between explanatory
variables accounted for
• Landscape-level
estimates of invasion
possible
Field Data
• Points located with sub-meter GPS
• 2 m diameter plots
390 on glacial landform
109 on debris flows
• Snow depth survey July 29- Aug 1,
2008
• All trees < 8m tall tallied by species
& height (1620 trees)
• 505 trees aged
Flow of Results
Spatial Patterns
• Snow depth in relation to
biophysical controls
• Tree abundance in relation to
biophysical controls
Temporal Patterns
• Tree invasion over time
• Tree invasion and climate
Interactions of Climate &
Biophysical Controls
Glacial Landforms
Results: Snow Depth & Biophysical Controls
Mean:
0.67m
95%CI:
0.6-0.8m
Debris Flows
More snow in depressions, lower elevations, distance from overstory
Nonlinear interactions between explanatory variables
More snow with less radiation, lower elevation, distance
More linear, reduced interactions, less variance described
Mean:
0.2m
95% CI:
0.1-0.3m
Results: Multi-Scale Controls of Snow
Landscape context is important
• Larger landforms
influence both overall
snow depth and micro site
controls of snow
• Smoother surface
on debris flows
• Greater wind
redistribution of snow on
smoother debris flows
Results: Tree Abundance & Biophysical Controls
• Mountain hemlock associated with microtopography and
distance to overstory canopy
• Silver fir strongly associated with distance to potential seed
sources, followed by microtopography
• Relationships between tree invasion and biophysical variables
much stronger on glacial landforms
Glacial landforms
Debris flow
Results: Temporal Patterns of Tree Invasion
• Not just an increase in densities
1950: 7.8% of meadow area with tree invasion
2008: 34.7% of meadow area with tree invasion
• Invasion dominated by Mountain hemlock, Pacific Silver Fir
restricted to under existing trees
• Invasion rate
greater on
debris flows
(0.75% v. 0.57% Yr-1)
Results: Climate, Landforms, & Invasion
Glacial Landforms
Debris Flows
• Annual snowfall
most important,
not temperature
• On debris flows
tree invasion not
associated with
annual snowfall
Adj R2 = 0.2887
p ≤ 0.01
Adj R2 = -0.0356
p = 0.5
• On debris flows,
positive
association with
spring snowfall!
Results: Climate, Microtopography, and Invasion
Only For Hemlock on Glacial
Landforms
• Hemlock invasion largely
in years with low snow on
ridgetops & midslopes
• During high snow years,
less invasion overall &
constrained to ridgetops
Conclusions
• Snow and tree invasion associated with multi-scale
landscape controls
• Species matter
• Landforms & topography alter both the spatial patterns of
tree invasion & response to climate
• Tree invasion on debris flow landforms
• Scale & landscape context matter
• Multiscale and context dependent responses pose
problems for modeling future responses to climate at
regional and global scales
• Need for experimentation (future climate now)
Acknowledgements
Funding provided by:
USDA Forest Service, Pacific Northwest Research Station
USDA Forest Service, Forest Inventory and Analysis Program
The Native Plant Society of Oregon
Hoener Memorial Fellowship, OSU
Waring Travel Grant, OSU
Thanks to field assistants:
Dan Irvine
Alex Gonsiewski
Questions?