The Hardy & Weinberg equilibrium principle If allele frequencies in a population are not fixed across
generations, the population is said to be evolving.
(1) In populations following the rules of Mendelian genetics,
allele frequencies do not change across generations
Evolution is defined as a change in allele
(2) If the two allele frequencies in a population are p and q, frequencies in a population across generations
then the genotype frequencies will be given by p2, 2pq, and q2
where:
How can we use the Hardy-Weinberg equilibrium principle?
In practice, if the allele frequencies of a population are found to
p2 (homozygote) + 2pq (heterozygote) + q2 (homozygote) =1 change from generation to generation, or if the genotype frequencies
(p + q = 1) cannot be predicted based on allele frequencies, then one or
more of Hardy-Weinberg model assumptions is being violated.
(3) The equation holds only for populations with: A change in allele frequencies allows
1. No selection; 2. No mutation; 3. No migration • To predict that evolutionary changes take place, and
4. No genetic drift; 5. And where individuals choose their • To assess which agents are responsible for the changes
mates at random
What happens to allele frequencies in a population in case Initial Frequency of Genotypes Genotypes New
that one of the Hardy-Weinberg assumptions is violated? Frequencies genotypes without with genotype
of alleles selection selection frequencies
Example: case of natural selection
(population
of 1000)
Assume B1 and B2 are two alleles of a gene, with frequencies of 360 All survive 360 0.45
B1 0.6 B1B1 0.36
0.6 and 0.4, respectively. After random mating under Hardy-Weinberg 480 75% “ 360 0.45
B2 0.4 B1B2 0.48
rules, the proportion of the genotypes is expected to be: 160 50% “ 80 0.10
B2B2 0.13
B1B1 (0.36), B1B2 (0.48), B2B2 (0.16). Given a population of 1000
zygotes, the number of individuals will be 360, 480, 160, respectively. Total survived: 800
New allele frequency after selection:
B1 = 0.45 + 1/2 x 0.45 = 0.675 (7.5 % increase)
B2 = 0.10 + 1/2 x 0.45 = 0.325 (7.5% decrease)
Now assume that individuals carrying the B2 allele are less favourable
survivors, and that: All B1B1 survive, 75% of B1B2 survive, and
50% of B2B2 survive. What happens to the proportion of genotypes
and alleles? Due to violation of the “no-selection” assumption the population
has evolved in response to selection.
How fast do allele frequencies change in response to selection? Case study: Evolution of the fruit-fly alcohol dehydrogenase
in response to selection
Cavener & Clegg (1981) worked on a population of fruit flies,
which have two alleles of alcohol dehydrogenase: ADHF (fast)
and ADHS (slow).
The fast enzyme breaks down ethanol at twice the rate of the slow
enzyme. Knowing that ethanol is poisonous, Cavener & Clegg
designed an experiment to study the ADH allele frequency in response
to selection.
They used molecular tools to test the presence of each allele
in response to selection
1
The experiment of Cavener and Clegg
Using molecular tools to identify alleles
C = control (no ethanol)
1. Enzyme mobility (protein analysis):
iso-enzymes (isozymes). Example of
alcohol dehydrogenase (ADH)
How would the graph
ADHF of the fast ADH allele
look like? Where does it start?
Where does it cross the graph
ADHS of the slow ADH allele?
What value will it reach?
F/S F/F S/S
E = with ethanol
2. DNA sequence
The control population (no selection) obeyed the Hardy-Weinberg rules with no
change in allele frequencies. The experimental populations, under selection
pressure, showed marked changes in allele frequencies.
Caution: Selection does not necessarily change allele frequencies
In spite of the strong selection against homozygotes,
the frequencies of the alleles have not changed:
Both alleles are hence the population has not evolved
equally affected
by selection. However, genotype & allele frequencies do not
Therefore no obey the Hardy-Weinberg equation: E.g. for
change in allele homozygotes the calculated genotype
frequencies. This frequency would be: 0.52 = 0.25, and for the
is only correct if heterozygote the calculated frequency:
both alleles have 2 x 0.5 x 0.5 = 0.5
the same (0.5)
frequency! In reality:
B1B1 genotype is 125/750= 0.166 < 0.25
B2B2 genotype is 125/750= 0.166 < 0.25
B1B2 genotype is 500/750= 0.666 > 0.50
p(B1) = (2 x 125 + 1 x 500): (2 x 750) = 0.5
p(B2) = (2 x 125 + 1 x 500): (2 x 750) = 0.5
Start here
Case of heterozygote superiority (over-dominance):
where heterozygotes have higher fitness than either homozygote
Hardy & Weinberg concluded that:
Hybrid vigour (heterosis): An example for
heterozygote superiority (over-dominance) (1) In populations following the rules of Mendelian genetics,
allele frequencies do not change
(2) If the two allele frequencies in a population are p and q,
then the genotype frequencies will be given by p2, 2pq, and q2
where:
p+q=1
p2 + 2pq + q2 = 1
Assumptions for a population obeying the
Hardy-Weinberg equilibrium principle:
1. No selection Evolution is the change in allele
Parent Parent Hybrid offspring 2. No mutation frequencies in a population across
The heterozygous F1 offspring is larger and stronger than either 3. No migration generations
homozygous parent 4. No genetic drift
5. Individuals choose their mates at random
2
Mutations can cause substantial
How do violations of the Hardy-Weinberg equilibrium change in allele frequencies, but very Start here
slowly. In the current example, it will take
principles affect allele frequencies and evolution? 1000 generations for allele A to
reach 0.8, and this assumes
a mutation rate of 1:10000,
Adding mutations to the Hardy-Weinberg analysis: How which is very high.
effective are mutations at changing allele frequencies
and as an evolutionary force?
As mentioned earlier, typical mutation rates are 10-4 to 10-9.
Consider a case with the high rate of 1:10,000 and see how it
affects allele frequencies.
Calculation of allele frequencies as a result of mutations:
A = 0.9 – [0.0001 x 0.9] = 0.89991
a = 0.1 + [0.0001 x 0.9] = 0.10009
Case study: Effect of mutations on fitness of bacteria
in response to selection
Lenski and co-workers tested the ability of bacteria to grow on a
Relative fitness
limiting nutrient medium for over 3000 generations (1500 days).
Samples of bacteria were collected each day and kept. At the end
of the experiment the relative fitness of the bacteria from each
generation was tested.
What is the meaning
of the “jump”?
Relative fitness = growth rate on limited medium
divided by growth rate on rich medium
Migration: (in evolution) the movement of alleles
between populations Adding migration
Mutations may have a fitness advantage, therefore selection will
ensure that these are maintained in the population (example to the Hardy-Weinberg
of sickle-cells). However, some mutations may remain in spite of analysis: example of
not having a fitness advantage. Why? the one-island model.
Because they are continuously created anew.
Starting point (example)
A2
A1 Genotype frequency on the
When the rate at which copies of a deleterious allele are
island is fixed at:
being eliminated by selection is exactly equal to the rate at which
A1A1 1.0
new copies are being created by mutation, the frequency of the
A1A2 0
allele is said to be at equilibrium, and the situation is called
A2A2 0
mutation-selection balance.
3
What happens on the island in case of migration The water snakes of Lake Erie
from the continent?
Virtually all banded snakes
Genotype
frequencies do
not obey Hardy- Mostly but not entirely
Weinberg rules unbanded snakes
Start here Virtually all banded snakes
Single-island model The model shows that gene flow from a continent to an island
will eventually drive the allele frequency on the island to a value
exactly equal to what it is on the continent.
Conclusion:
If migration is allowed to proceed unopposed by any other force of
evolution, it will eventually completely homogenize allele frequency
across populations.
If migration continues ‘for ever’ under the same conditions,
what would be the frequency of the banded allele on the island?
Genetic drift
Assume a pool of gametes with A1 or A2 alleles with the frequencies
Genetic Drift - Changes in the frequencies of alleles in of 0.6 and 0.4, respectively. According to the Hardy-Weinberg model
a population resulting from sampling error in drawing the zygote population should be 0.36 A1A1, 0.48 A1A2, 0.16 A2A2,
gametes from the gene pool to make zygotes, and from and the allele frequency should remain unchanged.
chance variation in the survival and/or reproductive success Now assume that we monitor each zygote as it is formed, one by one,
of individuals. and at each step we calculate the frequency of allele A1. If we
draw a graph of A1 frequency as a function of the cumulative number of
zygotes, how will this graph look like?
This is caused by small numbers of zygotes resulting in random
Frequency of allele A1
discrepancy between theoretical expectations and actual results. Discrepancy between theoretical expectations
1.0
The Hardy-Weinberg model is based on the assumption that and actual results in a small population
0.8
populations are large enough and does not apply when the
populations are small. 0.6
0.4
0.2
0
100 200 300 400
Cumulative number of zygotes made
4
In contrast to Natural Selection, which is a non-random mechanism Genetic drift (sampling error) as a mechanism of evolution:
of evolution, genetic drift is a mechanism of evolution that is If a few individuals colonize a new region, the resulting
absolutely random, which does not lead to adaptation, but does population will not have all the alleles found among
lead to changes in allele frequency. members of its source population
In the Hardy-Weinberg model, genetic drift results from violation
of the assumption of infinite population size. The founder event: The establishment of a new population,
usually by a small number of individuals
The founder hypothesis: The hypothesis that many
speciation events begin when small populations colonise
new geographic areas
The founder effect: A change in allele frequencies that occurs
after a founder event, due to genetic drift in the form of sampling
error in drawing founders from the source population
Principles of Natural selection - revisited
Natural selection cannot fashion perfect organisms
Natural selection acts on individuals, but its consequences occur
in populations. Evolution is limited by historical constraints, and does not build
complex structures from scratch.
Natural selection acts on phenotypes, but evolution consists of Example: As humans have developed from four-legged
changes in allele frequencies. ancestors, perfect adaptation for upright posture will be
Natural selection is a non-random process because it increases unlikely (and backproblems will never be abolished) .
adaptation to the environment.
Natural selection is not progressive in the sense of leading toward Extensive improvements of individual function generate new
a predetermined goal. There is no trend toward more advanced disadvantages – adaptations have to be compromises.
forms of life. Example: Humans have improved the flexibility of their hands
and limbs, which makes them more prone to sprains, torn
ligaments etc.
Combining Darwin’s theory of evolution with Mendelian genetics
Chance plays a greater role than originally expected, especially provides a comprehensive theory, with mechanisms and ability
when genetic drift singles out a random sub-population. to predict both qualitative and quantitative changes in populations.
Example: Individuals that migrate to a new location do not
necessarily contain the best alleles for adaptation to the new
environment. The Hardy-Weinberg model provides the basis for calculating changes
in populations, and for the development of complex models for making
intelligent prediction of populations.
New alleles do not arise on demand – evolution has to work with
the alleles that are available within a population.
Example: In the experiment that selected for the adaptation of
bacteria to growth on a limiting nutrient medium, a certain
improved fitness level could not be further increased.
5
What is the ‘individual’ unit of selection? The “Replicator” unit of selection (according to Richard Dawkins)
is defined as anything in the universe that causes copies
of itself to be made. According to Dawkins, a successful Replicator
Darwin regarded the individual organism as the unit of
should have the following three properties:
selection and this is certainly accepted by most scientists.
However, the debate about the unit of selection still goes on. Longevity: a Replicator has largely the same structure before and
after a copying or replication process.
Fecundity: A Replicator needs to make a certain number of copies to
Two main models exist in defining the properties of the persist as a Replicator. The more copies made by the Replicator, the
unit of selection: the “Replicator” and the “Interactor”. more successful it will be in out-replicating others.
Dawkins Vs. Gould: Survival of the fittest
Fidelity: A Replicator needs to have a certain degree of accuracy in
(2001; by K. Sterelny) Icon Books UK
its copies. However, it is copying errors (slight breakdowns in fidelity)
that drive evolution.
According to Dawkins and others, the gene has all that
is needs to be a true ‘Replicator’ and the actual unit of selection.
S. Gould and others reject the idea of the gene being the
unit of selection, claiming that although the marks of
The “Interactor” unit of selection has the following properties:
evolution are best kept at the genetic level, genes are not
causal units or agents of natural selection. An alternative
(a) Interacts with the environment, broadly construed as all surrounding
model is the “Interactor as the unit of selection”.
and influencing biotic and abiotic factors, in such a way that:
(b) One or more of their traits imparts differential reproductive success
through the interaction, so that:
(c) Relatively more or less (compared with other individuals at their
level) of their hereditary material (however packaged) passes to the next
generation.
Gould SJ & Lloyd EA (1999) Individuality and adaptation across levels
of selection: How shall we name and generalize the unit of Darwinism
PNAS 96: 11904-11909.