Upload
liliana-davalos
View
27
Download
2
Embed Size (px)
Citation preview
The importance of models —dynamics of extinction and their drivers from the Miocene to the last 40 years
Liliana M. Dávalos
National Science Foundation15 July 2016
!
Who am I? Evolution & Conservation
Our research mission
Biological diversityDiversification Extinction
Focus first part
Biological diversity
Colonization, speciation Extinctionincrease decrease
The debate: equilibrium
• Spp. richness ~ resource abundance
• Long-term spp. richness stable
• Diversity dynamics follow predictable course after perturbation
The debate: disequilibrium
• Communities are unsaturated
• Disturbance prevents communities from reaching equilibria
• Dispersal limitation slows reaching equilibrium
Elements of the equilibrium model MacArthur & Wilson 1963 Evolution
Most common application
Morgan & Woods 1986 Biol. J. Linnean Soc.
Extended application Dávalos & Russell 2012 Ecol. Evol.
Change in habitat ~ change in spp. Dávalos & Russell 2012 Ecol. Evol.
But larger Greater Antilles < spp. than expected Dávalos & Russell 2012 Ecol. Evol.
But what about the dynamics over time? Rabosky & Hurlbert 2015 Am. Nat.
Phylogenies and equilibrium
• Test of MW evolutionary equilibrium dynamics using phylogenies
Ricklefs & Bermingham 2001 Science
A model including speciation Valente et al. 2015 Ecol. Lett.
Key findings from model comparison
• Disequilibrium in most bird groups
• Tendency to equilibrium in Darwin’s finches
• Diversity-dependent
• Together, community dynamics are non-equilibrial
Valente et al. 2015 Ecol. Lett.
Rojas et al. 2016 Syst. Biol.
Rojas et al. 2016 Syst. Biol.
Figure 2
Ametrida Sphaeronycteris Centurio Pygoderma Ardops Ariteus Phyllops
† Phyllops silvai† Phyllops vetus
Stenoderma † Cubanycteris silvai
Artibeus (Artibeus)
† Artibeus anthonyiArtibeus (Koopmania)
Artibeus (Dermanura)
Ectophylla Enchisthenes Chiroderma Vampyressa (Vampyriscus)
Vampyressa (Metavampyressa)
Mesophylla
Vampyressa (Vampyressa)
Platyrrhinus Vampyrodes Uroderma
12 10 8 6 4 2 0
Mil l ion years before present
S A
M2.83
1.780.28
0.12
2.60
1.02
Founder-event dispersal(20% of cladogenetic events)
S A
M
1.13
0.32
6.465.04 0.68
0.73
Anagenetic dispersal(33% of cladogenetic events)
Miocene Pliocene
Tavares et al. in prep.
How old are the Greater Antilles? 35 Ma?
Iturralde-Vinent & MacPhee 1999 Bull. Am. Mus. Nat. Hist
Input data on Greater Antillean bats Valente, Etienne, Dávalos in review
Equilibrium dynamics! Valente, Etienne, Dávalos in review
Age Ma
λc μ γ λa
20 0.28 0.33 0.030 0.1945 0.32 0.41 0.036 0.16
What happens if no fossils? Valente, Etienne, Dávalos in review
Age Ma
λc μ γ λa
20 0.22 0.26 0.017 0.2745 0.26 0.35 0.021 0.27
Why are so many bats extinct? Climate? Humans?
It’s looking more and more like humans Stoetzel et al. 2016 Quat. Sci. Rev.
0 10 20 30 40
020
4060
80
Num
ber o
f end
emic
spe
cies
0 10 20 30 40
010
2030
4050
Num
ber o
f end
emic
spe
cies
Time (Ma)Time (Ma)
BA
How long to restore the species? Valente, Etienne, Dávalos in review
Conclusions part I
• First proposed in the 1960s, some evolutionary aspects of equilibrium theory can only be tested now
• Using phylogenies
• Longstanding questions remain to be answered, e.g.,
• is diversity predictable?
• relative contribution biotic vs. abiotic factors
Photo by Jon Flanders
More conclusions part I
• Need to include extinct species
• Underestimate extinction rate (obvious)
• Underestimate immigration rate (less obvious)
• Human effects large
• Long time to recovery
• Habitat change probably important Dávalos & Russell 2012 Ecol. Evol.
Focus second part
Biological diversity
Colonization, speciation
Land use changeincrease decrease
When the UN comes knocking The Special Session of 2016
But illicit crops are everywherePrevious focus on
Colombia
Why tropical deforestation? I
• 2 models
• Immiserization:
• poor growers bring marginal land into production
• external shocks (e.g., debt crisis)
• internal dynamics (e.g., demographic growth)
Rudel & Roper 1997 World Dev.
Why tropical deforestation? II
• Frontier:
• investors/government build roads
• roads bring colonization
• colonists clear land
• frontier stabilizes (development),
• or recedes (hollow frontier)
Rudel & Roper 1997 World Dev.
1400
1600
1800
2000
2200●
●
●
●
●●
●
●
3035404550
Forest
NP
1500
2000
2500
●
●
●
●
●●
●
●
28 30 32 34 36 38 40
Pasture
150200250300350400
●
●
●
●●
●
●
●
0.51.01.52.02.5
Coca
500
550
600
650
700 ●●
●
●●
●
●
●
3035404550
PA
(ha)
150
200
250
300
350
●●
●●
●
● ●
●
28 30 32 34 36 38 40
2468101214
●
●
●
●●●
●●
0.51.01.52.02.5
90
100
110
●
●
●
●
●
●
●
●
3035404550PLAND
EN
N (m
)
95100105110115120
●
●
●
●
●
●
● ●
28 30 32 34 36 38 40PLAND
300
400
500
600
●
●
●
●
●
●
●
●
0.51.01.52.02.5PLAND
Figure 2
BA
G
C
E FD
H I
253035404550
2001 2004 2007 201025
30
35
40
2001 2004 2007 2010
1
2
3
2001 2004 2007 2010
PLA
ND
K L M
The closing of the forest frontier
• An example from Guaviare, Colombia
• Forest declines, pastures increase
• Forest fragments become smaller and more isolated
• New stable equilibrium loses forest almost completely
Dávalos et al. 2014 Biol. Cons.
Where do illicit crops grow?
• In 2000s Colombia
• Correspond to forest frontier
• Interface between settled land and forests
• Spatially clustered
Dávalos et al. 2011 Environ. Sci. Technol.
What happened in the 1960s? Etter et al. 2006 J. Environ. Manage.
Opening the frontier Denevan 1966 Geogr. Rev.
Roads and colonization projects
Brücher 1977 Geogr. Z. Schuurman 1978 Tijdschr. Econ. Soc. Ge.
Brazil
Bolivia
Peru
Colombia
Ecuador
Coca cultivation
Government-sponsored
before 1979
1990/1992
Projects
Coca cultivation
High
Low
2014
Percent forest Probability of coca
a b c
A remarkable spatial pattern
Dávalos, Sanchez, Armenteras in revision
Brazil
Bolivia
Peru
Colombia
Ecuador
Coca cultivation
Government-sponsored
before 1979
1990/1992
Projects
Coca cultivation
Percent forest
a b
Location of projects ~ C emissions Harris et al. 2012 Science
Coca cultivation
High
Low
2014
Percent forestHigh
Low
Probability of coca
b c
Is coca spatially clustered around projects? Dávalos, Sanchez, Armenteras in revision
High
Low
Probability of coca
c
~40 years later, Amazonia coca remains clustered Dávalos, Sanchez, Armenteras in revision
Parameter 2.5 Mean 97.5Intercept -1.2671 -1.1540 -1.0410Distance to nearest project
-0.6096 -0.5613 -0.5129
Precision municipality intercept
1.116 1.4582 1.910
Precision spatial
0.243 0.5211 1.079
Coca is spatially associated with the frontier UNODC 2016
Can coca explain some deforestation?
• Three independent studies say no
• Dávalos et al. 2011 Environ. Sci. Technol. no effect after including social factors
• Armenteras et al. 2013 Reg. Environ. Change no effect of coca in Amazon
• Sánchez-Cuervo et al. 2013 Ecosystems no effect any time or place Photo by Ana Luz Porzecanski
No relationship coca cultivation ~ deforestation
ratesUnpublished
Conclusions part II
• Actions from 4 decades ago have large effects
• Brought colonists to the Amazon
• Farmers turned to coca
• Landscapes transform following roads
• End point often minimal forest
Photo by OSITRAN, Peru
More conclusions part II
• Spatial patterns misleading without history
• History illuminates the opening of the frontier
• Frontier deforestation can be connected to poverty
• But actually requires investment, e.g.,
• Roads
• Development projects
Thanks!