Evolutionary and Population Genetics I-III

Evolutionary Genetics
History: Mendelian Genetics

In his now famous experiments of the 1850s and 1860s
Gregor Mendel bred pea plants and examined the way that traits were passed
down across generations.
Mendelian Genetics
Mendel began 2 years of breeding experiments to determine
if his pea plants always bred “true”
that is, always produced a specific type of offspring
purple-flowered offspring when a purple-flowered

parent was self-fertilized
white-flowered offspring when a white-flowered

parent was self-fertilized.
Mendelian Genetics
In the parental generation
Mendel crossed a true-

breeding parent plant
homozygous for purple
flowers with a true-
breeding parent plant
homozygous for white
flowers.
Mendelian Genetics
Mendel then self-fertilized the F1

plants to produce an F2 generation
(the second generation of
offspring).
Mendelian Genetics
Mendel was able to derive a

number of important conclusions
about the genetics of diploid
organisms from these experiments.
Mendelian Genetics
From his experiments, Mendel
could infer the genetic
contributions of both parents to
their offspring.
He deduced that even though all F1 plants

produced purple flowers
they must have received and

retained genetic information from both
parents
Mendelian Genetics
From his experiments, Mendel
could infer the genetic
contributions of both parents to
their offspring.
Mendel’s results suggested
that each parent plant had two

copies of what we now call genes
that the two gene copies

separate with equal probability into
the gametes (sex cells) of the pea
plants.
Mendelian Genetics
We now know that this is correct.
Mendel’s first law, the law of

segregation, states that
each individual has two

gene copies at each locus
that these gene copies

segregate during gamete
production, so that only
one gene copy goes into
each gamete.
Mendelian Genetics
Mendel also concluded that the

effect of one allele can sometimes
mask the effect of another.
Mendelian Genetics
Mendelian Genetics
White flower color is recessive
It appears only when both gene

copies code for white flowers.
Darwin crafted his theory of evolution by natural selection
without having even a rudimentary understanding of the

mechanisms of inheritance.
One of Darwin’s greatest challenges was to explain how
sufficient variation was maintained in populations to
allow natural selection to continue to operate.
Not only does selection reduce variation by favoring

some forms over others
but in Darwin’s view, the very mechanism of

heredity also reduced variation.
Darwin, and most of his contemporaries, envisioned that
the process of heredity was a blending process
Under blending inheritance
the very process of heredity eliminates roughly half of
the heritable variation in a population every generation.
Mendel’s particulate theory of inheritance suggested that the
hereditary material
was more akin to colored

filters than to colored
liquids.
Phenotypic effects of these hereditary particles may blend
but the particles (what we call genes) remain distinct,
and can be separated again in future reproductive events.

Particulate inheritance explains not only why heredity itself does not
burn up variation
It also explains another dilemma associated with blending inheritance
When mixing occurs, how can new mutations ever have

significant effects on the characteristics of a population?
Natural selection could not increase or decrease the frequency of
this new mutation,
because the new allele is slowly “blending away.”

With Mendelian particulate inheritance
A new mutation would retain its particulate nature,
and not be blended into obscurity.
The Nature of Mutations
From an evolutionary perspective, the most important way to
categorize mutations is in terms of their effect on fitness.
With respect to a given environment, mutations can be
beneficial
deleterious
neutral
Today we know that there are no known mechanisms that
preferentially
generate mutations with a positive effect on fitness
avoid generating mutations with a negative effect on

fitness
The random nature of mutation was established through one of
the most elegant experiments in the history of biology.
In 1943, before geneticists knew for certain that DNA was

the hereditary material
Salvador Luria and Max Delbrück wanted to

understand the nature of the mutation process
Evolutionary biologists had proposed that mutations occurred

at random, independent of whether or not they would be
favored by natural selection
Luria and Delbrück knew that
when Escherichia coli was exposed
to a high density of a bacteriophage
—a virus that infects E. coli.
almost all of the E. coli cells

would be infected and killed.
But after some period of time,

colonies of E. coli that were
resistant to the phage would appear.
To explain this observation, Luria and Delbrück formulated
two alternative hypotheses:
Hypothesis 1: Random mutation
Prior to exposure to the phage, a few

resistant E. coli cells would arise by
random mutation.
Once exposed to the bacteriophage,

most cells would be killed, but the
resistant cells would not.
These would reproduce and form new

resistant colonies.
To explain this observation, Luria and Delbrück formulated
two alternative hypotheses:
Hypothesis 2: Acquired inherited resistance

At the time of exposure to the phage,
all E. coli cells would be sensitive to
the damaging effects of the phage.
The process of exposure to the phage

would induce mutations responsible
for phage resistance in a small fraction
of the bacterial cells.
the cells with induced resistance

would go on to produce colonies of
resistant cells.
To distinguish between these two alternatives, Luria and
Delbrück devised an ingenious experiment
They repeated the process with many cultures
…
From this information, they were able to distinguish between
the two hypotheses.
How?
Under the random mutation hypothesis
resistant cells that are present when the phage particles are added
must have had their origin in mutations that occurred earlier
during the growth of the bacterial population

If a mutation for resistance appears early in this growth process
it will become common in the population, giving rise to a large
cluster of colonies full of resistant individuals
Colony 1
= resistant cell Early Before phage

mutation present
= sensitive for
(not resistant) resistance
cell
If the first resistant mutation arises late in the growth process
it will generate a much smaller cluster of colonies full of

resistant individuals
Early Before Before
mutation phage phage
present present
Culture 1 for
resistance
Culture 2
Late
mutation
for
resistance
Under the acquired inherited resistance hypothesis, resistance never arises
until the phage particles are added.
At that point, each cell acquires resistance, or does not, independently from every other cell.
Because there are a relatively large number of cells in each culture
and a nontrivial fraction of these acquire resistance
by what is called the law of large numbers
each culture will have a similar number of resistant cells.
Before phage Before phage

present present
Culture 1 Culture 2
To distinguish between Before phage
present
Before phage
present
the two hypotheses,
Luria and Delbrück
carried out their protocol
repeatedly, and
they counted the

number of resistant Before phage Before phage
colonies that arose present present
from each of a large

number of cultures.
As predicted by the random
mutation hypothesis
Before phage Before phage
present present
Luria and Delbrück observed a

dramatic variation from culture to
culture in the number of resistant
colonies.
They concluded that phage

resistance was likely to be a
product of random mutations that
occurred at different times prior to Before phage
present
Before phage
present
the presence of the phage.
Types of Mutations
Refresher on some types of mutation:
Synonymous mutation (also called a silent mutation)
Nonsynonymous mutation (also called a sense mutation)
Nonsense mutation (also called a nonsense mutation)

The Frequency of Mutations
From an evolutionary perspective, the most important way to
categorize mutations is in terms of their effect on fitness.
With respect to a given environment, mutations can be
beneficial
deleterious
neutral
Mutation generates the genetic variation on which
natural selection acts.
If we want to understand how selection proceeds at the

molecular level
we need to know something about the rate at which

mutations occur
the distribution of beneficial, deleterious and neutral

mutations
Mutation rates vary considerably.
Among cellular organisms, mutation

rate per site increases with genome
size.
Mutation rates vary considerably.
Even among closely related animal species—primates, for

example—mutation rates can vary by an order of magnitude.
Mutations: Fitness Effects
Recently, evolutionary biologists have begun a concerted effort
to study the full spectrum of mutational effects, including
beneficial mutations.
Evolutionary biologists can estimate the distribution of

fitness effects in a number of different ways.
The most straightforward of these is to create an array of mutants
and then directly assess the fitness of each.
Recently, evolutionary biologists have begun a concerted effort
to study the full spectrum of mutational effects, including
beneficial mutations.
Peris and her colleagues experimentally

produced base substitutions at 100
nucleotide sites
Most mutations were either neutral or deleterious, though a
few were beneficial.
This distribution is much like that seen in a number of other viruses
We see the same sort of distribution in eukaryotes
Claudia Bank and her colleagues estimated the distribution of fitness effects
for mutations in the HSP90 heat shock protein of the yeast Saccharomyces
cerevisiae.
Generated an array of cells with different single-base mutations in the coding
region for this protein
At 36C, HSP90 is needed for thermal tolerance
As with viruses, most mutations were neutral or deleterious, though a

small number were beneficial
Population Genetics
What is Population Genetics?
Population genetics investigates
how the genotype frequencies in an
offspring population
are related to the genotype

frequencies in a parental
population.
To understand population genetics

work, we have to make a shift from
individual-level thinking
that is common in the study
of genetics.
Population Genetics
What is Population Genetics?
Population genetics investigates
how the genotype frequencies in an
offspring population
are related to the genotype

frequencies in a parental
population.
To understand population genetics

work, we have to make a shift from
individual-level thinking
that is common in the study
of genetics.
to the sort of population-level

thinking that is often
associated with ecology and
evolution.
Population Genetics
Genotype frequencies change over time
At its most basic level
Evolution occurs when

allele and genotype
frequencies change over
time.
Population genetics
provides a formal
structure with which to
look at this process.
Population Genetics
“Steady-state” frequencies
Population geneticists are not only interested in change.
They also want to understand stasis
when genotype frequencies or allele frequencies stay the same.
Are there “steady-state” frequencies for which no further change will occur?
Such frequencies are known as the equilibria.

Population Genetics
“Steady-state” frequencies
A system is at equilibrium
if the system has

reached a state
where it does not The marble here is in a state
change in the where it does not change
absence of outside position in the absence of
outside forces or processes
forces or processes acting on it.
acting on it.
Population Genetics
Hardy-Weinberg Model
Null models
If we want to understand the effects of

biological processes such as natural
selection or mutation on the frequencies
of genotypes in a population
we need a null model.
The Hardy–Weinberg model provides

such a null model.
Population Genetics
Assumptions of the Hardy-
Weinberg Model
As a null model, the Hardy–Weinberg model assumes that none of five
important evolutionary processes are operating:
1. Natural selection is not operating on the trait or traits in question.

2. Mating in the population is random with respect to the locus in
question. No assortative or disassortative mating.
3. No mutation is occurring.
4. There is no migration into or out of the population.
5. The population is effectively infinite in size, so genetic drift—
chance fluctuations in allele frequencies—is negligible.
Population Genetics
H-W Model
Every organism in our model population must have one of the three
possible genotypes
A1A1, A1A2, or A2A2.
Let’s call the frequencies of these three genotypes
f[A1A1]
f[A1A2]
f[A2A2]
Because each individual has one of these three genotypes,

the sum of genotype frequencies must be 1:
Population Genetics
H-W Model
From these genotype frequencies, we can compute the allele

frequencies directly.
Allele A1 is found only in individuals with the A1A1 or A1A2 genotypes.
Each A1A1 individual possesses two A1 alleles,

and
Each A1A2 individual possesses a single A1 allele
we can calculate the frequency of the A1 allele, which we

denote as p, from the genotype frequencies
Population Genetics
p = frequency of A1
Population Genetics
q = frequency of A2
Population Genetics
H-W Assumptions
because we have only two alleles in our system

it must be true that p+ q = 1
Population Genetics
To see how genotype frequencies change over time
we need to calculate the new genotype frequencies after individuals in our

population mate with one another and produce offspring.
Recall from our assumptions:

There is no assortative mating
mating in the population is random with respect to the locus in
question
Population Genetics
H-W Model
Population Genetics
H-W Assumptions
The offspring produced by random
draws from this gamete pool, occur
with frequencies that we can calculate
using the rules of probability:
Population Genetics
H-W Model
The frequency of the A1A1 genotype among the offspring is just

the frequency of the A1 allele, squared.
in the absence of evolutionary processes
the expected frequency of the A1A1 genotype

is equal to
the fraction of the time that we would expect

a random draw from a gamete pool with A1
at frequency p to yield two A1 alleles:
Population Genetics
H-W Model
The frequency of the A1A2 genotype is equal to the fraction of

the time that a random draw would select one A1 allele (p) and
one A2 allele (=q).
This is 2pq rather than pq

because there are two ways to draw an A1A2 individual
by drawing an A1 first and an A2 second
or by drawing an A2 first and an A1 second.
Population Genetics
H-W Model
The frequency of the A2A2 genotype is equal to
the fraction of the time that two A2 alleles would be drawn
q x q = q2
Population Genetics
H-W Model
Regardless of what p and q are initially
The Hardy–Weinberg model settles down to equilibrium

genotype frequencies of
p2, 2pq, and q2
after a single generation.

Population Genetics
Natural Selection
Recall Hardy–Weinberg assumes that none of five important

evolutionary processes are operating:

Population Genetics
Natural Selection
Population Genetics: Building Natural Selection into Our

Model
Begin by sketching out an example of natural selection that occurs in the

wild
then use the data from this example to make predictions about
allele frequency change.
Population Genetics
Natural Selection
Coat Color in Pocket Mice

Population Genetics
Natural Selection
Coat Color in Pocket Mice
Mice that live on light-

colored rocks tend to have
a sandy-colored coat
Mice that inhabit lava

fields are darker
Population Genetics
Natural Selection
The genetic control of coat color in rock pocket mice is also very similar
to the genetic control seen in oldfield mice.
In pocket mice, coat coloration is influenced by the same
melanocortin-1 receptor (Mc1R) gene that we saw in oldfield mice
In pocket mice, the Mc1R locus has two alleles -- A1 and A2
A1 allele is associated with dark coloration
A2 allele is associated with light coloration
A1 is dominant to A2
A1 A1 and A1 A2 individuals both display dark

coloration
Individuals with the A2 A2 genotype display light

coloration
Population Genetics
Natural Selection
We have a system in which an important character—coloration—is

associated with a single locus and clearly tied to survival.
Just how beneficial is it for an individual

to have the allele coding for a coat
coloration that matches the background
environment?
How advantageous is it for mice on the

dark lava fields to be A1 A1 or A1 A2?
How advantageous is it for mice on light-

colored rocks to be A2 A2 ?
Population Genetics
Natural Selection
To address these questions, Michael Nachman and his team collected

individuals at both lava sites and light-colored rock sites in an area
along the border between Arizona and Mexico
Population Genetics
Natural Selection
Focus on light-colored pocket mice living in the dark lava
Light-colored pocket mice living in the

dark lava fields suffered higher rates of
mortality.
Their chances of survival ranged from

60% to 98% of the chances of survival of
dark-colored mice on the dark lava fields.
Population Genetics
Natural Selection
Natural selection and pocket mice
Begin with the Hardy–Weinberg model
but we will relax Hardy–

Weinberg assumption about no
natural selection
allow natural selection to operate

on our population.
Population Genetics
Natural Selection
Label the frequency of the dominant

dark allele A1 as p.
The frequency of light allele A2 as q.

Population Genetics
Natural Selection
Natural selection and pocket mice
p = frequency of dark allele A1

Recall the A1A1 and A1A2 genotypes q = frequency of light allele A2
display dark coloration and are hard to
see against a dark lava field
whereas A2A2 individuals stand
out and suffer a reduced survival
probability.
On the lava fields, natural selection is thus

acting against the A2 allele.
Population Genetics
Natural Selection
To quantify the strength of natural selection against allele A2
p = frequency of dark allele A1

q = frequency of light allele A2
We use a parameter called the selection
coefficient, labeled s,
describes the fitness reduction of
the light phenotype relative to the

dark phenotype.
Population Genetics
Natural Selection
By convention, the fitness of the phenotype with p = frequency of dark allele A1

higher fitness—here the dark phenotype—is set to q = frequency of light allele A2
1.
The fitness of the other phenotype—here the light

phenotype—is set to 1−s.
s = 0 indicates no selection against an allele

s = 0.25 indicates a 25% reduction in fitness
s= 0.50 indicates a 50% reduction in fitness
…
Population Genetics
Natural Selection
For light-colored mice in dark lava p = frequency of dark allele A1

environments q = frequency of light allele A2
Nachman and his team
measured survival probabilities
ranging from 98% to 60% of that
experienced by the dark-colored
mice
exact value depended on the
population examined.
So, selection coefficients against light

coloration range from 0.02 (1- 0.98) to 0.40
(1-0.6).
Population Genetics
Natural Selection
A mathematical model that skips the math
Genotypes and fitnesses are p = frequency of dark allele A1

q = frequency of light allele A2
Population Genetics
Natural Selection
Genotype frequencies before and after selection:
H-W frequencies
Population Genetics
Natural Selection
Genotype frequencies before and after selection:
H-W frequencies
Multiply H-W
frequencies by
fitness
Population Genetics
Natural Selection
Sparing you the math
Genotype A1A1 A1A2 A2A2
Frequency before selection p2 2pq q2
Frequency after selection p2 2pq q2(1 ー s)
After lots of algebra

The frequency of p in the next generation
labelled p’ is equal to
p’ = pq2s/(1 – q2s)
Population Genetics
Natural Selection
p = the frequency of the A1 allele, the

dominant allele producing dark
phenotype
should increase by pq2s/(1 – q2s) in every

generation.
we can use this expression to p’ = pq2s/(1 – q2s)

plot the way that the allele
frequency of A1 would
change over evolutionary time for
different values of s.
Population Genetics
Natural Selection
Recall:
For light-colored mice in dark lava

environments
estimated selection coefficients

(s) against light coloration
ranged from 0.02 to 0.40. p’ = pq2s/(1 – q2s)
Population Genetics
Natural Selection
p’ = pq2s/(1 – q2s)
p = the frequency of the A1

allele, the dominant allele
producing dark phenotype
If s = 0.1,
When p, the frequency of the dominant dark allele, starts at a frequency of just 0.005 in
dark lava environments
we would expect that within 400 generations, it would increase to a frequency
near 1.
If s = 0.4 or 0.7,
the frequency of the dominant dark allele
would increase to a frequency near 1 even more quickly
Population Genetics
Natural Selection
p’ = pq2s/(1 – q2s)
p = the frequency of the A1

allele, the dominant allele
producing dark phenotype
Population Genetics
Natural Selection
Frequency-independent selection
The example of light and dark colored

mice
is a case of what is called
frequency-independent selection.
Frequency-independent selection occurs

when
the fitness associated with a
trait is not directly dependent on the
frequency of the trait in a
population.
Dark colored mice survive better on lava

fields independent of their frequency.
Population Genetics
Natural Selection
Frequency-dependent selection
Frequency-dependent selection occurs
when the costs and benefits associated with a trait depend on its
frequency in the population.
Frequency-dependent selection can be
positive
or
negative
Population Genetics
Natural Selection
Examples of positive frequency-dependent selection
Toxic species deter predators with

warning coloration
individuals benefit by being

recognized as inedible
natural selection should favor

individuals with the most common warning
colors
Population Genetics
Natural Selection
Examples of positive frequency-dependent selection
Oak trees often coordinate

the production of acorns
from year to year
producing many in a
“masting” year
few in other years.

Population Genetics
Natural Selection
Positive frequency-dependent selection
Under positive frequency-dependent selection
each phenotype is favored once it becomes sufficiently

common in the population.
If the phenotypes are controlled by two alternative alleles at a

single locus
one of the two alleles will eventually be fixed and the

other will be lost
Population Genetics
Natural Selection
1.0 0
Critical
Frequency of P2 phenotype threshold
Population Genetics
Natural Selection
Critical
threshold
Population Genetics
Natural Selection
Negative frequency-dependent selection
When negative frequency-dependent selection occurs
The fitness associated with a trait decreases as the frequency of the

trait increases in a population.
Each phenotype is favored when it is rare.
If the phenotypes are controlled by two alternative alleles at a

single locus, both alleles will be maintained in a balanced
polymorphism.
Population Genetics
Natural Selection
Researchers have found that some aspects of

feeding behavior in Drosophila larvae are under
genetic control.
Some larvae, known as “rovers,”

explore their environment widely
even when food is present
others, known as “sitters,”

exhibit limited movement as long
as they have something to eat
Population Genetics
Natural Selection
Both types are observed in natural populations of Drosophila

melanogaster.
Why hasn’t one type or the

other gone to fixation?
Population Genetics
Natural Selection
In a series of studies, Maria Sokolowski and her

colleagues demonstrated
that the sitter/rover dichotomy is due to a single

polymorphism at a locus dubbed forager.
Homozygotes for the recessive forS (forager-

sitter) allele
display sitter behavior
individuals with at least one copy of the dominant

forR (forager-rover) allele
exhibit rover behavior

Population Genetics
Natural Selection
Sokolowski and her colleagues

hypothesized negative frequency-
dependent selection was acting
foraging behavior.
Population Genetics
Natural Selection
To test this hypothesis,

the researchers allowed sitter
and rover individuals
to compete in a laboratory
environment in groups
composed of different phenotype
frequencies
Population Genetics
Natural Selection
Population Genetics
Finite Populations
Recall Hardy–Weinberg assumes that none of five important

evolutionary processes are operating:

Population Genetics
Finite Populations
So far, we assumed that populations are large
so large, in fact, that in every generation what is called

the law of large numbers applies to changes in gene frequencies.
The law of large numbers states that in

a large enough random sample
the realized frequencies
those frequencies we actually observe
will be very close to the expected
frequencies.
This is not what we expect when sample sizes are small.

Population Genetics
Finite Populations
Suppose that you tossed a fair coin 1,000 times

Analogy: population of 1000, two alleles, each at a frequency of 0.5.
At the end of the coin toss, the odds are that you would come very close to
seeing a 1:1 ratio of heads to tails.
If instead you only tossed your coin 10 times, you might get 5 heads and 5 tails
for another 1:1 ratio.
But over 75% of the time, you would get some other combination
4 heads and 6 tails, or 6 heads and 4 tails, or
3 heads and 7 tails, and so on.
Population Genetics
Finite Populations
In a smaller population, the law of large numbers will not hold
realized genotype frequencies may deviate substantially from the

expected genotype frequencies for any number of reasons.
Population Genetics
Genetic Drift
When allele frequencies change over time as a consequence of

chance alone,
these changes are said to be the result of

genetic drift
Population Genetics
Genetic Drift
Suppose that all our populations start with two alleles
A1 and A2 each at a frequency of 50%.
Further suppose
these alleles are selectively neutral

Population Genetics
Genetic Drift
10 different runs of the

simulation, with one
highlighted in blue for
visibility.
Population Genetics
Genetic Drift
Population Genetics
Genetic Drift
In a very small population

(100)
Frequency of A1 and A2
drift dramatically
As a result of genetic
drift, one particular allele
reaches a frequency of
100% in every one of
these populations.
When this happens, we
say that the remaining
allele has been fixed in
that population.
Population Genetics
Genetic Drift
For very small populations

100 generations is enough for one allele to go to fixation in every
run of the simulation.
For a population sizes of 100 and 1000, none of the simulations show an allele
going to fixation in 100 generations.
Population Genetics
Genetic Drift
Heterozygosity
Fixation of one allele means

Genetic drifts leads to marked decreases in heterozygosity
Genetic drifts leads to marked increases in homozygosity
Population Genetics
Genetic Drift
Though it is certain that some allele will eventually be fixed in each population in this model,
because genetic drift is a random process, it is not certain which allele it is
that will become fixed in which population.
Population Genetics
Genetic Drift
Mathematically, it can be shown that the probability that an allele at a neutral locus will
eventually be fixed
is equal to the initial frequency of that allele in the population.

Population Genetics
Genetic Drift and Founder Effects
One particular type of drift, known as the founder effect
change in allele frequencies that results from sampling effects

that occur
when small isolated populations, such as those on

small islands, are initially colonized (i.e. founded).
Population Genetics
Isabelle Gamache and her colleagues studied founder effects in the subartic black spruce (Picea
mariana)
Population Genetics
The genetics of dispersal and recolonization are particularly interesting in plants
Seeds have to be dispersed into a new area for the initial colonization to take place
but seed dispersal is not the only source of genetic variation for an establishing
population.
Pollen from other populations far away can blow in on the wind and fertilize the plants that
have become established there.
Population Genetics
Evolutionary biologists can tease apart the

patterns of pollen dispersal and the patterns of
seed dispersal:
Mitochondrial DNA is maternally inherited,

and thus is passed on only through seeds
mitochondrial DNA is absent in pollen.
The geographic distribution of mitochondrial

DNA should reflect patterns of seed dispersal
nuclear DNA will better reflect patterns

of pollen dispersal
Population Genetics
In an early study of black spruce in the Hudson Bay area, Gamache studied nuclear
DNA and found
no reduction of genetic diversity in post ice-age populations of black

spruce
no evidence of founder effects.
wind-dispersed pollen need not travel only from the leading-edge populations
during a recolonization event
Population Genetics
Gamache and her colleagues also examined the effect of migration via wind-
dispersed seeds by using mitochondrial DNA.
Founder Effects
The Case of the Black Spruce
Gamache and her team found that the migration of
mitochondrial DNA via seeds was
much more restricted and localized than the

migration of nuclear DNA via pollen dispersion.
The different types of nuclear DNA found in large parent

populations were represented in northern subpopulations.
When it came to mitochondrial DNA, although the southern

populations contained four different types of mtDNA
every one of the northern subpopulations had one and

only one type of mtDNA, called mitotype I.
Each of the Northern populations had fixed for one type of

mtDNA by random genetic drift. Mitochondrial DNA
Population Genetics
Natural Selection and Genetic Drift
“As soon as a gene is shown to have any effect whatever on fitness,

the conclusion is drawn that its distribution in populations must be
determined solely by selection and cannot be influenced by random
drift. But this is a logical non-sequitur."
Dobzhansky and Pavlovsky, 1957

Population Genetics
Theodosius Dobzhansky and Olga Pavlovsky examined the

relationship between drift and selection
As their trait of interest,

Dobzhansky and Pavlovsky
studied
naturally occurring genetic

chromosomal inversions in
the fruit fly Drosophila
pseudoobscura.
Population Genetics
Dobzhansky and Pavlovsky focused on two chromosomal inversions of the third
chromosome
These inversions are called the Pike’s Peak (P) and Arrowhead (A)
inversions.
Fitness-wise, A inversion > P inversion
Created twenty lines of fruit flies by crossing large numbers of

individuals from two wild populationsone from California and one
from Texas.
Fruit flies from the California population were

homozygous for P,
Fruit flies from the Texas population were homozygous

for A.
All individuals in each of the twenty experimental lines were

heterozygotes (P/A) at the third chromosome.
Genetic Drift & Natural Selection
All populations were

very large after 19
generations.
Any effect of drift

was due to initial
population size
differences across
treatments.
Population Genetics
Evidence of Natural Selection
Recall: Fitness-wise,
A inversion > P
inversion.
In both treatments,
the frequency of P
inversion decreased.
Population Genetics
Evidence of Genetic Drift
Greater variation in
end frequency of P
inversion when
initial population
size was large.
Population Genetics
In general, the interplay between selection and drift depends on the

strength of the selection and the population size.
When selection is strong and population size is large,

selection largely determines the change in allele frequencies.
When selection is weak and population size is small, drift

largely determines allele frequency change.
This relationship is captured by the rule of thumb that selection is

effective
when s >1/(2 x population size).

Evolutionary and Population Genetics I-III

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Evolutionary and Population Genetics I-III

Uploaded by

Copyright:

Available Formats

Evolutionary Genetics

History: Mendelian Genetics

purple-flowered offspring when a purple-flowered

white-flowered offspring when a white-flowered

Mendel crossed a true-

Mendel then self-fertilized the F1

Mendel was able to derive a

He deduced that even though all F1 plants

they must have received and

that each parent plant had two

that the two gene copies

Mendel’s first law, the law of

each individual has two

that these gene copies

Mendel also concluded that the

It appears only when both gene

without having even a rudimentary understanding of the

Not only does selection reduce variation by favoring

but in Darwin’s view, the very mechanism of

was more akin to colored

but the particles (what we call genes) remain distinct,

and can be separated again in future reproductive events.

When mixing occurs, how can new mutations ever have

because the new allele is slowly “blending away.”

With respect to a given environment, mutations can be

generate mutations with a positive effect on fitness

avoid generating mutations with a negative effect on

In 1943, before geneticists knew for certain that DNA was

Salvador Luria and Max Delbrück wanted to

Evolutionary biologists had proposed that mutations occurred

almost all of the E. coli cells

But after some period of time,

Hypothesis 1: Random mutation

Prior to exposure to the phage, a few

Once exposed to the bacteriophage,

These would reproduce and form new

Hypothesis 2: Acquired inherited resistance

The process of exposure to the phage

the cells with induced resistance

must have had their origin in mutations that occurred earlier

during the growth of the bacterial population

= resistant cell Early Before phage

it will generate a much smaller cluster of colonies full of

Before phage Before phage

they counted the

colonies that arose present present

from each of a large

Luria and Delbrück observed a

They concluded that phage

Synonymous mutation (also called a silent mutation)

Nonsynonymous mutation (also called a sense mutation)

Nonsense mutation (also called a nonsense mutation)

With respect to a given environment, mutations can be

If we want to understand how selection proceeds at the

we need to know something about the rate at which

the distribution of beneficial, deleterious and neutral

Among cellular organisms, mutation

Even among closely related animal species—primates, for

Evolutionary biologists can estimate the distribution of

Peris and her colleagues experimentally

At 36C, HSP90 is needed for thermal tolerance

As with viruses, most mutations were neutral or deleterious, though a

are related to the genotype

A2 allele is associated with light coloration