THEIS RESEARCH SEMINAR
Genetic adaptation of aspen populations
to spring risk environments:
a novel remote sensing approach
Haitao Li
Department of Renewable Resources
University of Alberta
Feb 16th , 2010
OUTLINE
• Background
o
o
o
Phenology & budbreak
Adaptation & genotype
Land surface phenology
• Methodology
o
o
Provenance trial
Remote sensing
• Results
o
Maps & graphs
• Discussion
Objective
• INVESTIGATION
Geographic patterns of genetic
variation in the spring phenology
of aspen
• INTERPRETATION
Adaptation of populations to
climatic risk environments
• IMPLEMENTATION
Seed transfer
• EXPLORATION
New method—remote sensing
Introduction
• PHENOLOGY
recurring biological phases of
species, biotic and abiotic
causes, and the
interrelation of phases within or
among species
• ADAPTATION
survival adaptation
capacity adaptation
• APPLICATION
Forest management, seed transfer,
climate modelling, …
Introduction
Spring phenology
Adaptive trait
Environmental contral
• HEATSUM
the thermal time, it is the
accumulation of degree-days
leading to a phenological
event
Degree-day :The average daily
temperature above a certain
threshold (0°C)
t1
Heatsum   GDD( xt )
t0
if
0
GDD( xt )  
 xt  Tb1 if
xt  Tb1
xt  Tb1
Heatsum
Temperature
• BUDBREAK
Threshold
Time
Introduction
• LAND SURFACE
PHENOLOGY
Forest canopy
spectral reflectivity
• MODIS
(Moderate Resolution Imaging
Spectroradiometer)
a key instrument aboard on
both Terra and Aqua
satellites lunched by NASA
Resolution
Spatial
250-1000 m
Spectral
0.62-14.38 µm
temporal
1-2 days
Introduction
• NDVI
Normalized
Difference
Vegetation Index
NDVI 
 nir   red
 nir   red
EVI  G 
 nir   red
 nir  (C1   red  C2   blue )  L
• EVI
Enhanced
Vegetation Index
Source Data from NASA:
http://honeybeenet.gsfc.nasa.gov/Sites/ScaleHiveSite.php?SiteID=MD001
Workflow of thesis
Climate data
Spatial pattern
No
Adaptation
Field
observation
Correlation
test
Spatial pattern
of genotypes
Spatial pattern
of genotypes
Yes
Remotely
sensed data
Field observation
• PROVENANCE
TRIAL
1998
43 open pollinated
bulked seed lots
Randomized complete
block design
Location of provenance trial
Research area and location of provenances
o43 provenances cover western Canada and Minnesota in US
o1126 trees in command garden
oobservation from May 4th to June 2nd 2009
Field observation
• 7-level bud
development scale
5
Index of Bud-break
0, dormant
1, buds swelling
2, buds breaking
3, extend to 1cm
4, extend to 2cm
5, extend to 3cm
6, fully extended
6
6
5
4
4
3
3
2
2
1
1
0
120
130
Days
140
Results of Field Observation
Symbols represent different ecological regions:
 BC taiga plains, AB northern boreal plains,  AB lower rocky mountain foothills,
 AB central boreal plains,  SK central boreal plains,  MN boreal shield.
Remote sensing methods
Greenup date
Map of green up date
From 2001 to 2005
EVI imageries
–MODIS EVI data from 2001 to 2005
–spatial resolution 500 meter
–16 days interval
Climate data
Average required heatsum
Plus standard error
Heatsum maps from 2001 to 2005
Mean temperature in the 130th day
2001
2004
2002
2005
2003
Average for five years
The distribution of Greenup date
2001
2004
2002
2005
2003
Five-year average
The distribution of Greenup date
2001
2004
2002
2005
2003
Five-year average
Heatsum requirements from provenance trial and remote sensing data
North
High Elev
Aspen Parklands
Validation
• Provenance trials
Similar spatial pattern for required heatsum
• Climate data
Storm-track from Rocky mountain to central Alberta
Correlation of heatsum with long-term
climate conditions
4
PCA: PC1 & PC2
MWMT
MN
AHM
SHM
MCMT
0
SK
FH
CT
EXT_Cold
NW
-2
BC
PAS
-4
Prin2
2
MAT
PCA (Red is high value and green is low)
-4
-2
0
2
Prin1
4
6
MCMT: mean coldest month temperature
SHM: summer heat/moisture index
PCA: PC3
AHM
0
1
FH
EXT_Cold MCMT
NW CT SKSHM
-1
MN
-2
PAS
-3
BC
PCA (Red is high value and green is low)
-4
Prin3
MWMT
-4
-2
0
Prin2
2
4
AHM: annual heat/moisture index
MCMT: mean coldest month temperature
MWMT: mean warmest month temperature
PAS: precipitation as snow
Correlation of heatsum with long-term
climate conditions
 Precipitation and dryness in
winter and early spring
affect the timing of budbreak
Interpretation
• Survival adaptation
vs.
• capacity adaptation
o North & high elevation
short growing season
break early
o Dry areas
aspen parklands and other
place
breaking bud in late spring
too dry to grow
Growth traits: Height and DBH
Graphed by Tim Gylander
Growth traits: Productivity
Mapped by Pei-yu Chen
Application
X
What’s neW
Environmental
control
• Budbreak delayed by
doughtiness and frost
damage
Technology
• Detecting genotype by
remote sensing
Acknowledgement
• Andreas Hamann, Xianli Wang, Elisabeth Beubia,
Pei-yu Chen, Nicholas Coops, Arthuro Sanche
• NSERC, Alberta-Pacific Forest Industries,
Ainsworth Engineered Canada LP, DaishowaMarubeni International Ltd., Western Boreal
Aspen Corporation, Weyerhaeuser Company Ltd.