Specialized Herbivore
Feeding Leads to Increased
Speciation in Plants
Shuya Kyu
Gregg Hartvigsen
A Lovely Honey Locust
www.washington.edu/ home/treetour/hlocust.html
1
www.washington.edu/ home/treetour/hlocust.html
Poisonous Milkweed
http://www.laspilitas.com/butterflies/Butterflies_and_Moths/Monarch/Monarch_butterfly.htm
2
http://www.laspilitas.com/butterflies/Butterflies_and_Moths/Monarch/Monarch_butterfly.htm
Outline
1.
2.
3.
4.
The Model
Methods
Results
Conclusions
3
Outline
1.
2.
3.
4.
The Model
Methods
Results
Conclusions
The Model
32 × 32 toroidal grid
Stationary plants
4
The Model
32 × 32 toroidal grid
Stationary plants
Herbivores randomly walk 5 steps
The Model
32 × 32 toroidal grid
Stationary plants
Herbivores randomly walk 5 steps
5
The Model
At the end of a generation…
• Organisms mate with their neighbors.
• Leave offspring depending on their success.
• Discrete generations.
The Model
What determines the success of an individual?
31-bit binary haploid
Genesgenome
(Interaction Code)
6
The Model
Feeding
100% match
50% match
The Model
Feeding
Herbivores increase Biomass
Plants decrease Biomass
Individuals with large biomass are preferably
chosen as mates and leave more offspring
7
The Model
Mating
Rule: Individuals have to be within
2 mismatches to mate
The Model
Mating
Crossing Over
Mutation
Progeny
8
The Model
3500
Number of Individuals
3000
2500
2000
1500
1000
500
0
10
81
161
401 481 561
801 881 961
200 241 321 400
600641 721 800
1000
Generations
The Model
• Saves your time
• Complete information
3500
Number of Individuals
3000
2500
2000
1500
1000
500
0
0
1
81
200
1000
800
161 241 321 401 481 561 641 721 801 881 961
400
600
Generations
9
The Model
Do the plants and herbivores influence the
evolution and speciation of each other?
The Model
Population as a graph
10
The Model
Population as a graph
1. Each individual is a vertex
2. Individuals are connected if their
genotypes are within 2 mismatches
01111011
01110100
00110100
11110100
11111111
00100100
00101100
00110111
The Model
Population as a graph
Define a connected component as a species
01110100
01111011
00110100
11110100
11111111
00100100
00101100
00110111
11
The Model
Population as a graph
Properties of the dominant species
The Model
Population as a graph
Properties of the dominant species
• Rate of Evolution
12
The Model
Population as a graph
Properties of the dominant species
• Rate of Evolution
n
Rate of Evolution =
Σ
( Ci(t)-Ci(t-1) )
i=1
Generation
The Model
Rate of Evolution
01
11
(0, 0.66)
Center of Mass
(0.17, 0.5)
Rate of Evolution = 0.33
00
10
13
The Model
Population as a graph
Properties of the dominant species
• Characteristic Path Length
The average number of steps between
any two vertices in a graph
The Model
Feeding Function
0.5
Specialist
Proportion of
the plant that 0.33
can be eaten
Mc = Critical
number of
Generalist
mismatches
20
30
Number of Mismatches
14
Outline
1.
2.
3.
4.
The Model
Methods
Results
Conclusions
Methods
0.5
Proportion of
the plant that
can be eaten
6 Feeding Settings
0.33
20
30
Number of Mismatches
15
Methods
3500
Transient
Phase
Number of Individuals
3000
Plants
2500
Herbivores
2000
1500
1000
500
0
10
81
161
401 481 561
801 881 1000
961
200 241 321 400
600641 721 800
Generations
Outline
1.
2.
3.
4.
The Model
Methods
Results
Conclusions
16
Results
1. Number of species
2. Rate of evolution
3. Characteristic Path Length
Results
1. Number of Species
Number of Species
12
10
Specialist
a
Generalist
a
b
8
Plants
b,c
6
4
p
Herbivores
c
A
2
A,B
B
B
B
d
B
d
0
20
22
24
26
28
Mismatches allowed for feeding
30
Just
JP
Plants
p < 0.001 for plants and herbivores
17
Results
Hamming distance per time step
2. Rate of Evolution
Specialist
Generalist
0.8
b
0.7
0.6
0.5
b
a,b
b
a,b
Plants
a,c
A
0.4
Herbivores
A,B
0.3
B,C
0.2
C
C
c
C
0.1
0
20
22
24
26
28
Just
JP
Plants
30
Mismatches allow ed for feeding
p < 0.001 for herbivores, p = 0.003 for plants
Results
3. Characteristic Path Length
Characteristic Path Length
3
2.5
2
Specialist
a,b
A
Generalist
a
a
A
a
a
b
B
1.5
C
Plants
Herbivores
D
C,D
c
1
0.5
0
20
22
24
26
28
30
Mismatches allow ed for feeding
Just
JP
Plants
p < 0.001 for plants and herbivores
18
Outline
1.
2.
3.
4.
The Model
Methods
Results
Conclusions
Conclusions
1. The number of plant species increases as the
herbivore feeding becomes
more specific
p
Number of Species
12
Specialist
Generalist
10
8
Plants
6
Herbivores
4
2
0
20
22
24
26
28
30
JP
Mismatches allowed for feeding
19
Conclusions
2. The specialist herbivores evolve faster than
the generalist herbivores
Hamming Distance / Time
Rate of Evolution
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
20
22
24
26
28
30
Herbivory Specificity (# m ism atches allow ed)
Specialist
Generalist
Conclusions
2. The specialist herbivores evolve faster than
the generalist herbivores
Hamming Distance / Time
Rate of Evolution
Specialist
0.4
0.35
0.3
0.25
0.2
0.15
0.1
0.05
0
Generalist
20
22
24
26
28
30
Herbivory Specificity (# m ismatches allow ed)
Specialist
Generalist
A more adapted individual has a greater relative
advantage in the specialist herbivore setting
20
Conclusions
3. The Characteristic Path Length is positively
correlated with the Rate of Evolution
Characteristic Path Length
Rate of Evolution
Plants
Herbivores
Plants
Herbivores
Conclusions
3. The Characteristic Path Length is positively
correlated with the Rate of Evolution
Herbivore
Attack
Herbivore
Attack
21
Acknowledgements
•
•
•
•
•
•
•
Gregg Hartvigsen
Chris Leary
Tony Macula
Gary Towsley
SUNY Geneseo Biomath Initiative
The Natural Science Foundation
The Geneseo Foundation
22