Transcript Locatelli_2008_Migration

Seminar series on climate change and forest management, BC Ministry of Forests and Range
Future Forest Ecosystem Initiative.
November 6, 2008. Teleconference between Canada and Costa Rica
Effects of landscape connectivity on
ecosystem adaptation to climate
change in Central America
Bruno Locatelli, CIRAD-CIFOR, Indonesia
Pablo Imbach, CATIE, Costa Rica
EuropeAid/ENV/2004-81719
1
Climate change and ecosystem migration
• CC modifies location of suitable environments for plants
(Hannah et al., 2002)
• Future distribution of ecosystems depends on the ability
of plants to migrate (Pitelka et al., 1997; Kirilenko et al., 2000)
– High migration rates (>1000 m/yr) required to adapt to CC
(Malcolm et al., 2002)
• Many studies on CC and ecosystems consider unlimited
dispersal or no dispersal (Pearson, 2006)
• Uncertainties on migration rates of trees:
– During post-glacial warming
• 100’s to 1000’s m/yr
(Pearson 2006, Malcolm et al., 2002),
• less than 100 m/yr (McLachlan et al., 2005)
– Models of current migration
• Around 100 m/yr (Dyer, 1995)
2
Biological corridors and protected areas
• Landscape alteration may reduce
migration capacity (Pitelka, 1997)
– Altering dispersal rate
– Reducing suitable habitat for successful colonization
• Corridors can enhance landscape connectivity between valuable
vegetation areas (e.g. protected areas)
• Their role in migration under CC depends on spatial patterns
Protected
Area
Corridor
Direction of species
movement due to CC
• Latitudinal and altitudinal gradients
3
– Is it so simple?
Can corridors facilitate ecosystem migration
between protected areas in a context of CC?
Case of the Mesoamerican Biological Corridor
in Costa Rica
4
Study site
• Costa Rica
– Biological richness & conservation policies
– “Increasing isolation of protected areas may
prevent them from functioning as an
effective network” (Sánchez-Azofeifa et al., 2003)
Mexico
Belize
Guatemala Honduras
El Salvador Nicaragua
Costa Rica
Panama
Protected Areas
Biological Corridors
• The Mesoamerican
Biological Corridor (MBC)
5
(CCAD-UNDP/GEF, 2002)
– Regional initiative
– Conceptualized and agreed
upon in 1997
– Under progressive
implementation
Nicaragua
N
Costa Rica
Protected Areas
Mesoamerican
corridor
Pacific Ocean
6
Panama
Caribbean
Sea
Approach and model assumptions (1/2)
• Cellular Automata
Holdridge
• Simple representation of
vegetation
– Model ≠ evolution of vegetation
– Model = landscape connectivity
– Vegetation types =
Holdridge life zones
– Each vegetation type is composed of 5 groups of species with
different migration rates
• From slow (100 m/yr) to fast-moving (2000 m/yr)
7
Model assumptions (2/2)
• Migration is possible through contiguous
pixels with vegetation
• 4 scenarios
– No migration
– Migration
• With vegetation only in
protected areas
• With vegetation in protected
areas and corridors
• With vegetation everywhere
(“Eden”)
8
Algorithm
Controller
Pixel px
Initialize
Run
Estimate 1990 Life
Zone (lz1990)
Create baseline presence:
presence(px,spft,lz1990)=1
Estimate Current
Life Zone (lzcurrent )
sp adapted to lzcurrent?
No: presence:=presence/2
Yes: sp already present in px?
Yes: presence:=presence*2
Next
decade
(with upper bound
due to carrying capacity)
Update
indicators
9
Species spft,lz
No: sp in neighbourhood?
(radius=migration(ft))
Yes: presence:=0.25
No: presence:=0
Data
• Scales
– Spatial = 2.5 arc min (around 4.6 km x 4.6 km)
– Temporal = Decades, from 1990 to 2050
Nicaragua
Nicaragua
Costa Rica
Caribbean
Sea
Costa Rica
Caribbean
Sea
Pacific Ocean
Pacific Ocean
Panama
• Climate and altitude
– WorldClim dataset (Hijmans et al., 2005):
• Altitude, Monthly Precipitation and
Temperature
– Current and future
• IPCC scenario A2
• HadCM3 climate model
10
Protected Areas
Biological Corridors
Panama
• Protected Areas and
Biological Corridors
– CCAD map, 2001
Results and discussion
12°N
Life zone
movement
8°N
11
86°W
81.5°W
Protected areas most sensitive to CC
(no migration scenario)
Nicaragua
N
Caribbean
Sea
Pacific Ocean
High impacts
Low impacts
12
Panama
Vegetation Index (Relative to Eden Scenario)
Protected areas under different
migration scenarios
Evolution of Vegetation in Protected Areas
1
0.9
Effect of
corridors
0.8
0.7
0.6
0.5
1990
Protected areas and corridors
Protected areas only
No migration
2000
2010
2020
Years
13
2030
2040
2050
Which protected areas benefit more
from corridors?
14
Which corridors contribute more
to the adaptation of protected areas?
15
How to explain these results?
Missing
corridors?
Corridors and Areas not evaluated
Direction of Life Zone Movement
Important
corridors
(connecting
isolated and
vulnerable
areas, in a
good
direction)
16
Already connected
protected areas
Conclusion
• Sensitivity of protected areas to CC
– Especially isolated areas and in mountain and dry zones
• Role of corridors for reducing vulnerability
• Useful tool for:
– Identifying vulnerable protected areas
– Prioritizing corridors in a context of CC
• Further steps:
– Consider actual vegetation outside
protected areas and corridors
– Use different climate scenarios
– Use different representations of
ecosystem dynamics and migration
– Perform sensitivity analysis
17
Thank you!
18