Download The evolution of Lake Erie watersnake color

Detecting Natural Selection in Real Time:
Examples from Lake Erie Snake Populations
Rich King
Department of Biological Sciences
What is Natural
Selection?
As many more individuals of each
species are born than can possibly
survive; and as, consequently, there
is a frequently recurring struggle
for existence, it follows that any
being, if it vary however slightly in
any manner profitable to itself,
under the complex and sometimes
varying conditions of life, will have
a better chance of surviving, and
thus be naturally selected. From
the strong principle of inheritance,
any selected variety will tend to
propagate its new and modified
form.
What is Natural
Selection?
As many more individuals of each
species are born than can possibly
survive; and as, consequently, there
is a frequently recurring struggle
for existence, it follows that any
being, if it vary however slightly in
any manner profitable to itself,
under the complex and sometimes
varying conditions of life, will have
a better chance of surviving, and
thus be naturally selected. From
the strong principle of inheritance,
any selected variety will tend to
propagate its new and modified
form.
Variation
What is Natural
Selection?
As many more individuals of each
species are born than can possibly
survive; and as, consequently, there
is a frequently recurring struggle
for existence, it follows that any
being, if it vary however slightly in
any manner profitable to itself,
under the complex and sometimes
varying conditions of life, will have
a better chance of surviving, and
thus be naturally selected. From
the strong principle of inheritance,
any selected variety will tend to
propagate its new and modified
form.
Variation
+
Fitness
Differences
What is Natural
Selection?
As many more individuals of each
species are born than can possibly
survive; and as, consequently, there
is a frequently recurring struggle
for existence, it follows that any
being, if it vary however slightly in
any manner profitable to itself,
under the complex and sometimes
varying conditions of life, will have
a better chance of surviving, and
thus be naturally selected. From
the strong principle of inheritance,
any selected variety will tend to
propagate its new and modified
form.
Variation
+
Fitness
Differences
+
Inheritance
What is Natural
Selection?
As many more individuals of each
species are born than can possibly
survive; and as, consequently, there
is a frequently recurring struggle
for existence, it follows that any
being, if it vary however slightly in
any manner profitable to itself,
under the complex and sometimes
varying conditions of life, will have
a better chance of surviving, and
thus be naturally selected. From the
strong principle of inheritance, any
selected variety will tend to
propagate its new and modified
form.
Variation
+
Fitness
Differences
+
Inheritance
Evolution
How can natural selection be detected?
Darwin’s view:
“. . . natural selection will always act with extreme slowness . .
.”
“Its action depends on . . . physical changes, which are
generally very slow . . .”
“. . . variation itself is apparently always a very slow process.”
“The process will often be greatly retarded by free
intercrossing.”
“. . . I do believe that natural selection will always act very
slowly, often only at long intervals of time.”
By extension:
Natural selection should be difficult to detect
How can natural selection be detected?
A more contemporary view (J. A. Endler, 1986, Natural Selection in the Wild):
“Natural selection is ubiquitous enough to be found in a wide
variety of organisms . . .”
“. . . strong selection is by no means uncommon in natural
populations.”
By extension:
Natural selection should be readily detected by any of a
variety of methods.
I. Correlation with environmental
factors
II. Comparisons between closely
related species
III. Comparisons between unrelated
species living in similar habitats
IV. Deviation from formal null models
V. Long-term studies of trait frequency
distributions
VI. Perturbation of natural populations
VII. Genetic demography or cohort
analysis
VIII. Comparisons among age classes or
life-history stages
IX. Nonequilibrium prediction of
fitness differences or their
consequences
X. Equilibrium prediction of the
outcome of natural selection
Western Lake Erie
Islands
• 18 islands
• 1 – 4000 hectares
• 1 – 30 km from mainland
• shorelines of limestone and
dolomite that resisted glaciation
• isolated ca. 4000 years ago
• islands = natural laboratories
of evolutionary biology
• renowned for abundance of
snakes: “Les Isles aux
Serpentes”
Lake Erie
Watersnakes
• Medium-sized nonvenomous live-bearing
colubrid
• Active April – October
• Courtship May – June
• 20+ offspring in
September
• Maturity in 2 – 4 years
• Longevity 10 + years
• Maximum size > 1 m
• Aquatic prey
• Terrestrial
retreats/hibernation
Variation in Lake Erie Watersnake Color Pattern
Mainland
populations
Island
populations
Variation in Lake Erie Watersnake Color Pattern
Mainland
populations
Island
populations
Inheritance of Lake
Erie Watersnake Color
Pattern
• Pattern elements remain fixed
in shape and position
• Resemblance among relatives
suggests significant heritability
(h2)
• Captive matings suggest a
major locus with regular
dominant to reduced pattern
h2 = 0.20 - 0.34
h2 = 0.40 - 0.62
Evidence for Natural Selection
Method I: Correlation with environmental factors
Environmental differences impose different selective regimes, resulting
in differences in phenotype among populations
Island shoreline
Mainland marsh
Evidence for Natural Selection
Method I: Correlation with environmental factors
Environmental differences impose different selective regimes, resulting
in differences in phenotype among populations
Island shoreline
Mainland marsh
Suggests color pattern is target & predation is agent of selection
• Other targets & agents?
• Current vs. past selection?
• Strength of selection?
Method IX: Nonequilibrium prediction of fitness
differences or their consequences
Phenotype
Performance
Do snakes differing in color
pattern differ in crypsis (the
degree to which they are visible
to predators)?
Selection
Island
shoreline
Mainland
marsh
Island
shoreline
Mainland
marsh
Quantitatification of
crypsis: compare size
of pattern elements
between snakes and
backgrounds
Method IX: Nonequilibrium prediction of fitness
differences or their consequences
Phenotype
Performance
Selection
Performance functions predict that on islands, visual predators
should impose selection favoring less patterned morphs among
neonates and more patterned morphs among adults.
• Current vs. past selection?
• Strength of selection?
Method VII: Genetic demography or cohort
analysis (longitudinal analysis)
Do snakes differing in color pattern differ in survival?
• 317 newborns
marked &
released in fall,
1983
• 54 survivors
recaptured in
spring &
summer, 1984
• color-pattern of
survivors was
compared to
snakes that were
not recaptured
Method VII: Genetic demography or cohort
analysis (longitudinal analysis)
Do snakes differing in color pattern differ in survival?
Yes, among newborns, survival was highest for snakes with
reduced patterns.
• Color pattern is target of selection
• Survival (via differential predation?) is agent of selection
• Selection is occurring currently
• Selection differential () = -0.30 for number of dorsal
blotches
• Relative fitness of regularly patterned snakes is 78-90% that
of reduced pattern snakes (s = 0.1 – 0.22).
Method VIII: Comparisons among age classes
or life-history stages (cross-sectional analysis)
Selection should result in
differences in color pattern
frequencies between different
age classes.
• Prediction from crypsis
analysis (Method IX) is that
frequency of regularly
patterned snakes should be
greater among adults than
subadults.
• Compared color pattern
frequencies between
subadults and adults at 5
island sites.
Method VIII: Comparisons among age classes
or life-history stages (cross-sectional analysis)
Selection should result in differences in color pattern frequencies
between different age classes.
No consistent pattern of
selection from subadults to adults;
selection differentials
() = -0.05 – 0.01.
•Adult snakes have
outgrown most
significant predators?
Method V: Long-term studies of trait frequency
distributions
Natural selection should result in trait frequencies that remain constant
over time (stabilizing selection, directional selection balanced by gene
flow) or change monotonically (directional selection).
Morph frequencies
have remained
relatively constant
from 1980 – present
Method V: Long-term studies of trait frequency
distributions
Natural selection should result in trait frequencies that remain constant
over time (stabilizing selection, directional selection balanced by gene
flow) or change monotonically (directional selection).
Morph frequencies
have changed
markedly between
historic (<1961) and
recent (1980-2003)
samples
• Change in
selection?
• Change in gene
flow?
Method IV: Deviation
from formal null
models
In the absence of selection,
variation in color pattern
frequencies among populations
should equal that resulting
from the combined effects of
random genetic drift and gene
flow.
• Assessed allozyme
(protein) variation at 5
island and 2 mainland sites
(presumably neutral)
• Compared to variation in
color pattern allele
frequency
Method IV: Deviation from formal null models
• FST provides a measure of genetic differentiation among
populations
• Variation in putative color pattern allele is similar to that of
allozymes among 5 islands
• Variation in putative color pattern allele is markedly greater than
that of allozymes when mainland populations are included
5 Islands
5 Island + 2
Mainland sites
Allozymes
0.016
0.074
Color pattern
0.019
0.493
• Pattern is consistent with selection favoring different color pattern
alleles in island vs. mainland populations
Method X: Equilibrium prediction of the outcome
of natural selection
Color pattern variation in Lake
Erie watersnakes appears to
represents a balance between
1) selection favoring a
reduction in color pattern in
island populations and
2) gene flow from the mainland
where regularly patterned
snakes predominate.
Information on inheritance,
strength of selection, and
rate of gene flow can be
used to predict equilibrium
conditions.
Method X: Equilibrium prediction of the outcome
of natural selection
Under an island-continent model with selection favoring reduced pattern
snakes on islands and gene flow from mainland populations consisting
solely of patterned snakes, change in allele frequency,
q =
q – hq2
- q - mq
1 – hq2
Where
q = frequency of reduced pattern allele
h = the strength of selection = 1-(1/(1-s))
m = rate of gene flow.
Using allozyme-based estimates of FST and Nm and markrecapture estimates of N, m ≈ 0.0008
From cohort analysis (Method VII), s ≈ 0.10
Predicted equilibrium
allele frequency:
q = 0
Observed allele frequencies are lower than predicted
• Gene flow underestimated?
• Strength of selection overestimated?
Predicted equilibrium
allele frequency:
q = 0
Observed allele
frequencies
Conclusions
Multiple methods demonstrate convincingly that Lake Erie
watersnakes are subject to natural selection
1) Color pattern is the target of selection
2) Survival (via differential predation) is the agent of selection
3) Selection occurs in present-day populations
4) Strength of selection is moderately high
Some unexpected results and directions for future research
1) Predicted pattern of selection on older snakes (from analysis
of crypsis) was not evident
2) Historic and recent samples exhibit unexplained differences
in morph frequency
3) Observed color pattern allele frequencies differ from
predicted equilibrium allele frequency
4) Selection in mainland habitats has not received detailed
study
More generally: watersnakes are not unique
Endler’s 1986 compilation documents selection in 141 species
involving 314 traits and includes 566 estimates of selection
coefficients & gradients (Methods VI, VII, VIII)
Kingsolver et al. summarize >2,500 estimates of selection
coefficients and gradients from 1984-1997(Kingsolver et. al. 2001,
American Naturalist 157:245-2261; Hoekstra et al. 2001, PNAS 98:9157-9160)
Examples include morphological, physiological, biochemical, and
life history traits and mortality, fecundity, and sexual selection
Examples of directional selection outnumber stabilizing selection
Shortcomings (there is more work to be done!)
Lack of replication, small sample size
Morphological traits and directional selection predominate
Viability, fecundity, and sexual selection need to be studied on
similar timescales
More sophisticated analyses (fitness surfaces, population
genomics) are warranted