Upload
victor-norton
View
220
Download
0
Tags:
Embed Size (px)
Citation preview
Remote Sensing for Biodiversity Conservation, Land cover and Land use Change and Carbon/Ecosystem
Management
Catherine Graham Stony Brook University
(many contributions – individual slides)
Remote Sensing for Biodiversity Conservation, Land cover and Land use Change and Carbon/Ecosystem
Management
Catherine Graham Stony Brook University
(many contributions – individual slides)
Improving assessment and modelling of climate change impacts on global terrestrial biodiversity
– McMahon et al. 2011
• Critical challenges were presented at the IPCC Working Group 2 (2007) – still many gaps in knowledge remain.
• “In common with other areas of global change science, the credibility of these predictions is limited by incomplete theoretical understanding, predictive tools that are acknowledged to be imperfect, and insufficient data to test, develop and improve model predictions.”
• What are these gaps? and How is NASA science filling them?
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Modified (slightly) from McMahone et al. 2011 Trends in Ecology and Evolution
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Monitoring programs• Remote-sensing• Biological data•Phenology •Rates
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Species’ ability to adapt•Genetic variation•Phenotypic plasticity•Migration
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Range models (species/functional group)•Correlative•Physiological•Population dynamics
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Community structure and dynamics• Species interactions –(disease, competition)•Food webs
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Integrative models• Biogeochemical models• Extinction risk models• Invasive/disease species spread models• Changes in distribution of species and functional groups • Influence of disturbance (disease/fire) on productivity
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Monitoring programs• Remote-sensing• Biological data•Phenology •Rates
Are ocean deserts getting larger?
Irwin and Olivier. 2009. Geophysical Research Letters.
RS data used:SeaWiFS/AVHRR
!! !! !!! !! !!
!!
!!!!!!!!!!!!!!!!!!
!!!! !!!
! !!!!!
!!!!!!
!! !! !!!
! !!
!!!
!! !!!!!!!!!!!!!!
!!!! !
!! !
!!!! !!
!
! !!
!!!! !!!! !!!!!!!
!!!!!!!!! !!
!!!!
! !!!! !!
! !! !!!!! !!!! !! ! !!!!!!
!!
!! !! !!!
!!
! !!!!!!!!!!!!!
!
!
!!!
!
! !! !! !!
!
!
!
!
!
!!! !!
!
!
!
!
! Survey routes
Study sites
Forested ecoregions
km1000±
Disturbance and bird biodiversity (BBS data)- Forest harvest
Rittenhouse et al. 2010 PLoS
Landsat used to quantify land cover change1985-2006
Current and past forest disturbances affect progressive similarity of forest birdsProgressive similarity - community similarity for each subsequent year relative to the baseline
All forest birds Midstory and canopy Neotropical migrants GroundTemperate migrants CavityPermanent residents Interior forest
Rittenhouse et al. 2010 PLoS
Gaps in our knowledge of global ant diversity
Lots of ant data
Not so many data
No-analogueclimates
Jenkins et al. (2011) Diversity
and Distributions.
Predicted Future Ant Diversity
Generalized Linear ModelClimate: temperature, precipitation, aridityGeography: biogeographic regionInteractions: region * climate
Jenkins et al. (2011) Diversity and Distributions.
No-analogueclimates
16Nemani et al., 2003, EOM White & Nemani, 2004, CJRS
TOPS: Common Modeling Framework
Monitoring, modeling, and forecasting at multiple scales
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Species’ ability to adapt•Genetic variation•Phenotypic plasticity•Migration
Genetic and morphological variation across taxa mapped using RS data (MODIS products, Q-scat)
Red – genetic diversity
Blue – morphological diversity
Yellow - bothThomassen et al. 2011
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Range models (species/functional group)•Correlative•Physiological•Population dynamics
Manderson, Palamara, Kohut , Oliver in press. Marine Ecology Progress Series
Sea surface temp Divergence, HF radar
Dynamic layers
Climate model
Static layers
Current occurrences
Future projected species habitat (time series of maps)
Current environmental conditions
Projected future conditions1. 2
.3.
4.
2100
2010
SDM
Velasquez, Salaman and Graham
More Andean bird species are predicted to loose habitat than to gain it with climate
COLONIZATIONS LOSSES
RS data used:MODIS productsQ-Scat
Distribution of Antarctic and sub- Antarctic penguin colonies
Rapid warming
Olivier and colleagues
Significant Changes in Ideal Breeding
Habitats: 1978-2010
Chinstrap Habitats
Adelie Habitats
Gentoo Habitats
Olivier and colleagues
Changes in penguin habitat suitability correspond to empirical changes in abundance of penguins at the Palmer Station,
Antarctica
Changes in habitat suitability within 75 km of Palmer Station.
Percent change in population trends from initial sampling (Ducklow et al. 2007)
Can richness be monitored and forecasted?
Coops, Waring, Wulder, Pidgeon and Radeloff. 2010. Journal of biogeography
Based on the annual sum, the minimum, and the seasonal variation in monthly photosynthetically active radiation, fPAR from MODIS
Dynamic Habitat Index
Woodland bird species richness can be predicted by the Dynamic Habitat Index
Dynamic habitat index can be used to forecast patterns of species richnessof woodland/forest birds.
Coops, Waring, Wulder, Pidgeon and Radeloff. 2010. Journal of biogeography
OBSER VED
PREDICTED
Broad scale estimates of forest bird species richness are consistent across studies
Models derived from BBSRS data – Lidar canopy
structure predictor variables, mODIS
Goetz et al. (forthcoming) Global Ecology & Biogeography
Lidar used to map multi-year prevalence / optimal breeding habitat..
Black throatedblue warbler
Goetz et al. (2010) Ecology 91:1569-1576
Hubbard Brook Experimental Forest
Habitat groupDeciduous, evergreen forest(2001 NLCD)
ConstraintsEdge & area sensitivityForest composition (FIA)Housing density
Intrinsic elementsSnags/logsUnderstory vegetationForage/prey abundance
Main modeling unit; general habitat requirements
Species-specificmodifiers
Habitat needs not mapped at large spatial scales; need to be maintained within each habitat group
Beaudry et al. 2010 Biological Conservation
Building potential habitat models using nested habitat elements
Anna M. Pidgeon, Fred Beaudry, Volker C. Radeloff
Habitat groupDeciduous, evergreen forest(2001 NLCD)
ConstraintsEdge & area sensitivityForest composition (FIA)Housing density
Intrinsic elementsSnags/logsUnderstory vegetationForage/prey abundance
Main modeling unit; general habitat requirements
Species-specificmodifiers
Habitat needs not mapped at large spatial scales; need to be maintained within each habitat group
Beaudry et al. 2010 Biological Conservation
Building potential habitat models using nested habitat elements
Anna M. Pidgeon, Fred Beaudry, Volker C. Radeloff
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Range models (species/functional group)•Correlative•Physiological•Population dynamics
Linking environmental data to physiological response over large scales
Kearney, Simpson, Raubenheimer and Helmuth 2010, PTRS
• Biophysical (Heat Budget)
Model
• Dynamic Energy Budget
Model
• Growth, reproduction,
size
•Environmental data
• Survival, distribution
20
30
40
50
60
70
80
90
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
snou
t-ve
nt le
ngth
(cm
)
years
Max of SVL
05
10152025303540
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
wet
mas
s (g)
years
Min of Mass
HeatDeath
ColdDeath
Egg?
More accurate predictions are made when daily remote-sensing data are used in models
0-50% shade, 10cm burrow
0
2000
4000
6000
8000
10000
12000
14000
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5re
serv
e de
nsity
(g/c
m3 )
years
Reserve
MaxRes
0
2000
4000
6000
8000
10000
12000
14000
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
rese
rve
dens
ity (g
/cm
3 )
years
Reserve
MaxRes
20
30
40
50
60
70
80
90
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
snou
t-ve
nt le
ngth
(cm
)
years
Max of SVL
05
10152025303540
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
wet
mas
s (g)
years
Max of Mass
HeatDeath
ColdDeath
Egg?
monthly data daily datasize
reserve
mass/repro (8 clutches)
size
reserve
mass/repro (11 clutches)
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Range models (species/functional group)•Correlative•Physiological•Population dynamics
Predicting Extinction Risks under Climate ChangePredicting Extinction Risks under Climate ChangeDynamic
layersClimate model
Static layers
210020
10SDM
2010
2100
Metapopulation model with dynamic spatial structure6.
Demographic model
000
000
000
3
2
1
44332211
S
S
S
SmSmSmSm
5.
Extinction risk assessment7.Synthesis across species to inform IUCN Red List process
8.
Akçakaya & Pearson
Predicting Extinction Risks under Climate ChangePredicting Extinction Risks under Climate ChangeDynamic
layersClimate model
Static layers
210020
10SDM
Metapopulation model with dynamic spatial structure6.
Demographic model
000
000
000
3
2
1
44332211
S
S
S
SmSmSmSm
5.
Extinction risk assessment7.Synthesis across species to inform IUCN Red List process
8.
Akçakaya & Pearson
2010
2100
Predicting Extinction Risks under Climate ChangePredicting Extinction Risks under Climate ChangeDynamic
layersClimate model
Static layers
210020
10SDM
2010
2100
Metapopulation model with dynamic spatial structure6.
Demographic model
000
000
000
3
2
1
44332211
S
S
S
SmSmSmSm
5.
Extinction risk assessment7.Synthesis across species to inform IUCN Red List process
8.
Akçakaya & Pearson
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Community structure and dynamics• Species interactions –(disease, competition)•Food webs•Guild/functional group structure
Phytoplankton diversity from ocean color
• Phytoplankton class-specific approach used in conjunction with SeaWiFS 10-year time series of surface Chl data in the global ocean
• Microphytoplankton (mostly diatoms) are major contributors in temperate-subpolar regions (50%) and coastal upwellings (70%) during the spring-summer season
• Nanophytoplankton (mainly prymnesiophytes) provide substantial ubiquitous contribution (30–60%)
• The contribution of picophytoplankton reaches maximum values (45%) in subtropical oligotrophic gyres
Contribution (%) to total primary production in boreal summer
Stramski and colleagues
Models accurately predict change of ecosystem engineershindcasts of limits (lines) and observed historical limits (dots), geographic region in grey
Predicting satellite derived patterns of large-scale disturbances in forests of the Pacific Northwest region response to recent climate variation(Waring, Coops and Running)
Physiologically informed models of 15 species of conifers
Physiological models and remote-sensing provide similar insights into ecosystem function
Stress of species predicted using a physiological informed models corresponds to areas that Disturbance predicted using physiological basis
Physiological models and RS measures provide the same pattern in Leaf Area Index (correlated maximum growth potential)
Land surface temperature & EVIMildrexler et al. 2009
Proportion of species stressed between 2005-2009 compared to baseline conditions (1950-1975)
~70% variation explained
Monitoring programs
Species’ ability to adapt
Range models
Community structure and dynamics
CURRENT BIOLOGY FORCASTING
Integrative models
ECOSYSTEM MANAGEMENT
Modified (slightly) from McMahone et al. 2011 Trends in Ecology and Evolution
What next?
• Linking RS time-series data biological data to better predict future biological diversity– Key for decision making– Key for inputs into biogeochemical models
• Determining what RS data captures in terms of biological diversity or ecosystem stress