Transcript The Evolutionary Significance of Chance: Mating Systems

Evolutionary significance of
stochastic forces and small
populations
Coyne JA, Barton NH and Turelli M. 1997. A
critique of Sewall Wright’s shifting balance theory
of evolution. Evolution 51:643-671
Genetic differentiation
• Evidence for population differentiation in
plants is indisputable.
– Deterministic forces (Natural selection)
– Stochastic processes (Genetic drift)
Drift causes random changes in allele frequencies
Simulated population; N = 10
Determinants of drift
• small population size (N)
• restricted dispersal (m)
N
N
N
N
population
N
neighbourhood
Effective population size, Ne
- a standardized measure of population size
- size of an ‘idealized’ population with the same
strength of genetic drift as the target population.
- the census number (N), adjusted for skewed sex
ratio, perenniality, selfing, persistent seed bank,
ploidy, non-random variation in fecundity etc.
- most cases, Ne is less than the actual count of
individuals in the population (N)
How important is chance?
• Darwin (1859): acknowledged that historical
accidents and chance could oppose the forces of
natural selection
• Gulick (1872): Hawaiian land snails
• Hagedoorn, A. L. and Hagedoom, A. C.
The Relative Value of the Processes
Causing Evolution. Pp. 294. Martinus
Nijhoff. The Hague, 1921.
Wright and Fisher
Fisher: adaptive evolution results simply
from Darwinian mass selection.
Wright: adaptation cannot be explained by
selection alone. Stochastic processes such
as genetic drift often play an important role.
Shifting Balance Theory
Fitness
Fitness landscape
selection
selection
drift
Genotype/phenotype
Coyne, Barton and Turelli 1997
“….it seems unreasonable to consider the
shifting balance process as an important
explanation for the evolution of adaptation”
Role of small populations and genetic
drift in the evolution of mating systems
in Eichhornia paniculata
Eichhornia paniculata
•Pontederiaceae
•short-lived
perennial/annual
•insect pollinated
Ephemeral water bodies
in Brazil, Cuba, Jamaica,
parts of Central America
Tristyly
•3 mating types
• mating is disassortative and
outcrossing
• stable state: frequencydependent selection
maintains equal morphs
frequencies
N = 167 populations
Estimate mating
type frequencies
Trimorphic = 118
Dimorphic = 42
Monomorphic = 7
Mating type structure
•Trimorphic populations near
1:1:1, or low on S
•Most dimorphic pops missing
the S morph;
•All monomorphic pops are M
How is mating system measured?
1. Need 8-10 half sib
offspring from each of 2030 mothers
3. Infer the genetic
contribution of the
paternal parent
Mother = AA
AB
AA?
AB
AB
2. Genotype mothers and
offspring using genetic
markers (allozymes,
microsatellites, AFLPs)
4. Estimate the rate of
outcrossing (t) that
produces the distribution
of offspring observed. S =
1-t
Population outcrossing rate varies
with mating type diversity
Cross-fertilizing
3-mating types
Self-fertilizing
1 mating type
Selfing variant of the M morph
What evolutionary forces have caused the
the loss of mating types and the transition
from a stable outcrossing breeding
system to self-fertilization?
Hypotheses
• Natural selection against the S morph,
perhaps related to pollinator x mating type
interactions
• Stochastic events associated with small, shortlived populations
Selection
Fruit set (% of flowers)
• Pollinator limitation: long-tongued solitary bees; may be
unpredictable in small pops; S morphs may be most
vulnerable
Fertility in the field
100
but S < M,L in 3 of 6 pops
F = 0.31, p > 0.50
90
80
70
60
50
40
30
20
10
0
L
M
Style Morph
S
Effective Population Size (Ne)
•Individual-based
simulations of
tristylous populations
• When Ne < 40, drift
can overcome
selection and cause
the loss of mating
types.
• Ne < 10, more likely
to lose two mating
types.
Mating types not lost equally
S morph - most likely to be lost
• frequency-dependent
selection resists loss of
morphs
•if 1:1:1, all morphs equally
likely to disappear due to
sampling error
• however, S allele is only
carried by S morphs and
thus cannot segregate from
remaining L and M.
ssmm
ssMm
SsMm
ssMM
SsMM
SSMm
SSMM
Effective population size in 10 populations
of E. paniculata
Genetic method
Sample allele frequencies
over at least 2 years

1 
p(1 p)[1 1
 ]
 2N e 
Variance in allele freq.
t
V(p) =

Ne
Ne - Demographic method
Five estimates
1. Estimate # of individuals
2. N, corrected for variation in among
years
3. N, corrected for variance in flower
production
4. N, corrected for mating type frequency
5. N, corrected for self-fertilization
Ne
Mean N = 763 (range 30.5 - 5040)
Mean Ne = 15.8 (range 3.4 - 70.6) Mean Ne / N = 0.106
Ne < 40 in 120 of 167 pops
160
Ne/N Demography
Temporal var = 0.47
Reprod effort = 0.42
Selfing rate = 0.98
Morph freq = 0.95
Ne-demographic
140
120
100
80
60
40
20
0
0
20
40
Ne-genetic
60
80
Effect of drift on
Spatial variation in morph structure
Predictions
Effective population size
Spatial variation in mating type structure
Dimorphic/monomorphic
Trimorphic
Temporal variation in frequency of S mating type
S morph lost from pops
Change in morph diversity / yr
0.5
Obs
Exp
0.4
0.3
0.2
0.1
0
1-25
26-50
51-150
151-500
Population size (N)
501-1500
Temporal variation in S as a function of N
What accounts for the loss of the L morph?
• Reproductive assurance: ability to self-fertilize in the
absence of pollinators favours selfing M morph
F=2.8, p = 0.13
Fruit set (% of flowers)
100
90
80
70
60
50
40
30
20
10
0
L
M
Style Morph
Why doesn’t the M morph spread in
trimorphic populations?
• pollinators not
scarce in large
pops
• siring
advantage
doesn’t exist
when S is
present
Fitness landscape
selfing
Fitness
outcrossing
selection
selection
drift
Genotype/phenotype