Variation in a population

1
Selection and Genetic Variation
1) selection against recessive alleles   If
alleles are recessive lethal, then selection
can only act on them when they are
homozygous   consider Dawsons flour beetles
started with population of all
heterozygotes, / l l / l is lethal, but / l
is same as wildtype /
2
Selection and Genetic Variation
1) selection against recessive
alleles   Although selection initially removed
the l allele from population at a rapid rate,
with each generation the frequency of l
declined more slowly
3
Selection and Genetic Variation
2) selection against homozygotes   This
population was started with 100 heterozygotes
for a viable allele V, and an allele L that is
lethal when homozygous although selection
rapidly caused the V allele to increase in
frequency, the L allele never disappeared in
fact, the frequency of L stabilized at 0.21  
4
Selection and Genetic Variation
2) selection against homozygotes   1/5th of the
population carried the lethal allele at
equilibrium (the point where the population
ceased to evolve) Why?  
5
Selection and Genetic Variation
3) selection against heterozygotes  
consider the case of flies with compound
chromosomes
normal pair of homologous chromosomes
compound chromosomes arms swapped - one ends
up with both left halves - other ends up with
both right halves
when these flies make sperm/eggs, meiosis gets
screwed up... they make 4 kinds of gametes
6
Selection and Genetic Variation
3) selection against heterozygotes   -
Flies can be homozygous for C (compound) or N
(normal) allele - two N/N flies can reproduce
all zygotes are viable (fitness 1) - two C/C
flies can reproduce 1/4th of zygotes viable
(fitness 0.25) - C/N flies dont exist they
never develop (fitness 0)
C and N flies cant make viable zygotes
together
7
Selection and Genetic Variation
3) selection against heterozygotes   one or
the other allele quickly becomes fixed in a mixed
population  
8
Selection and Genetic Variation
3) selection against heterozygotes   one or
the other allele quickly becomes fixed in a mixed
population - why? if there are few N/N flies,
the odds of 2 mating are low - most N/N flies
will not produce viable offspring - the allele
will vanish - if there are many N/N flies, they
quickly out-breed C/C flies, due to their
4-fold advantage in producing viable
offspring this is underdominance
9
Models of heterozygote superiority and
inferiority
- in overdominance (heterozygote fitness gt
homozygote fitness), population fitness is
maximized at its stable internal equilibrium,
the point to which the population naturally
returns
10
Models of heterozygote superiority and
inferiority
- in underdominance (homozygote fitness gt
heterozygote fitness), the population fitness
is minimized at the unstable internal
equilibrium, the point from which the population
naturally diverges
11
Frequency-dependent selection
Attack other fish by sneaking up, rushing
them, biting off a mouthful of scales -
Those with mouths that curve to the right
attack the left side of victims, and
vice-versa - Handedness of mouth is
determined by a single locus with 2 alleles
(simplest case!) - Right-handedness is
dominant
scale-eating fish of Lake Tanganyika
12
Frequency-dependent selection
- victims come to expect attacks from the
direction that the majority of the
scale-eaters attack from, at that particular
time - when right-handed fish are more common,
victims pay less attention to their right
side (where few attacks come from) this
gives left-handed fish the edge - as
left-handers get more food, they survive and
reproduce better - then, when left-handed
offspring are the majority, the situation
reverses
13
Frequency-dependent selection
proportion of left-handers
- squares proportion of successful breeding
adults
14
Frequency-dependent selection
proportion of left-handers
15
Frequency-dependent selection
The equilibrium point should be 50/50 of each
phenotype so what are the expected allele
genotype frequencies?   Alleles R L Allele
frequencies 0.3 0.7   Possible
genotypes RR RL LL   Hardy-Weinberg
predicts R2 2RL L2   Genotype
frequencies 0.09 0.42 0.49
16
Frequency-dependent selection 2
Another case pea aphid Acyrthosiphon pisum
occurs in green and red color morphs - what
maintains polymorphism, the occurrence of
both phenotypes in the population? Different
ial vulnerability to predation versus parasitism,
depending on color - green aphids are more
parasitized by wasps that lay their eggs
inside aphids - red aphids get eaten more by
ladybugs (theyre more obvious sitting there
on green plants)

17
Mutation as an evolutionary force
Mutation is ultimately responsible for creating
new alleles and genes, but..   - can mutation
also represent an evolutionary force, by
changing allele frequencies? - can mutation
affect the predictions of Hardy-Weinberg
equilibrium?  
18
Mutation as an evolutionary force
Consider a population where allele frequencies
are   A a (a recessive, loss-of-function
allele) 0.9 0.1   In the ordinary
Hardy-Weinberg state, adult genotypes will
be   AA Aa aa 0.81 0.18 0.01  
19
Mutation as an evolutionary force
Now assume A mutates to a at a rate of 1 per
10,000 genes each generation   due to
mutation, the allelic makeup of gametes will
be   A a 0.9 (0.9)(0.0001) 0.1
(0.9)(0.0001) 0.899991 0.10009  
20
Mutation as an evolutionary force
When gametes randomly fuse to form zygotes, the
genotype frequencies will be  
AA Aa aa 0.80998 0.18016 0.01002 Hardly
any change mutation had little effect over one
generation Over thousands of
generations, mutation can affect allele
frequencies 
21
Mutation as an evolutionary force
Alleles may be kept in a population through a
balance between mutation (creating deleterious
alleles) and selection (removing them) -
in mutation-selection balance, the frequency with
which new alleles are created by mutation
equals the rate at which they are
eliminated by selection When the frequency of a
harmful allele (say, cystic fibrosis) is
higher in a population than youd expect from the
mutation rate of that gene, then you have
reason to suspect some other force (i.e.,
selection) may be keeping that allele around
22
Mutation as an evolutionary force
Why does the ?F508 allele, which causes cystic
fibrosis, occur at a high frequency (0.02) in
populations of European descent? - selection
against homozygotes is strong - mutation rate is
too low to explain high allele frequency
23
Mutation as an evolutionary force
More importantly, mutation promotes evolutionary
change by genetic innovation - once a rare
beneficial allele is created by mutation, it can
rapidly become fixed in the population
through selective sweeps
bacteria evolved in a series of jumps
24
Migration
Migration is the movement of alleles between
populations   Migration can rapidly change allele
frequencies, especially for small
populations - individuals leaving a continent
make little difference to the allele
frequencies on that continent - those arriving
on an island with a small population can
make a huge difference to allele frequencies on
the island
25
Migration
Example banded vs unbanded water snakes
26
Migration
Example banded vs unbanded water snakes -
one gene w/ 2 alleles determines banded, unbanded
or intermediate morph - natural
selection favors banded snakes on mainland,
where they are cryptic (hidden from
predators) - selection favors unbanded
snakes on islands, where bands stand out
when snakes sun themselves on rocks to
warm up
27
distribution of banded vs unbanded snakes
28
Migration
Why doesnt selection fix the unbanded allele on
islands? (drive it to a frequency of 100)
- migrants from mainland continually introduce
banded allele into island population
- about 13 snakes per year move to islands, which
have 1300 snakes (roughly 1 migration
per year) Migration acts as a homogenizing
force - equalizes allele frequencies among
populations makes them more similar than
they would otherwise be
29
Genetic Drift
A sampling process (flipping a coin, drawing
beans from a bag) may produce results different
from theoretical expectations - flip a coin
four times, and you may get 4 heads   When the
actual results differ from theory, this is
sampling error Sampling error depends largely
on the number of samples drawn - flip a coin 40
times, and you are very unlikely to get 40
heads - will probably get 20 heads, give or
take a few
30
Genetic Drift
Sampling error in production of offspring in a
population is genetic drift Initial frequencies
are always heavily skewed during random
sampling - ie, drawing alleles one at a time
from a big batch ( gene pool)
31
Genetic Drift
Sampling error is very sensitive to population
size - as population increases, effects of
genetic drift diminish
- odds of getting the expected allele
frequencies when you make 10 zygotes by
drawing alleles at random
32
Random fixation of alleles
pop. size 4
Given enough time, any allele will eventually
become fixed or disappear if genetic drift is
the only mechanism at work - when one allele
is fixed, all others have a frequency of
zero - the odds that any given allele will
be the one that goes to fixation is the
initial frequency of that allele
40
400
33
Genetic Drift
(1) Every population follows a unique
evolutionary trajectory, because sampling
error affects allele frequencies at random - if
selection were at work, different populations
would evolve along similar trajectories
(2) drift works faster and stronger in small
populations - allele frequencies change more
dramatically if population size is small
(3) even in large populations, drift can cause
substantial evolution over long times -
geographic isolation results in differentiated
populations
34
Genetic Drift
(1) Every population follows a unique
evolutionary trajectory, because sampling
error affects allele frequencies at random - if
selection were at work, different populations
would evolve along similar trajectories
(2) drift works faster and stronger in small
populations - allele frequencies change more
dramatically if population size is
small Question to ponder What forces prevent
drift from fixing alleles in natural
populations?
35
Random fixation and loss of heterozygosity
Frequency of heterozygotes decreases over time,
as alleles drift towards fixation or
extinction - all else being equal (no
selection, etc), the frequency of
heterozygotes should fall in every generation -
given by the relationship Hg1 Hg 1 - 1
where N is population size
2N You are trying to maintain a group of 50
endangered llamas despite your
efforts to arrange random matings...
36
Random fixation and loss of heterozygosity
Heterozygosity decreases in every generation,
but more slowly in large populations - the
faster an allele disappears due to drift, the
more quickly you lose heterozygosity What
can reverse this effect?
37
Random fixation and loss of heterozygosity
tested experimentally by Buri with fruit
flies   - started 107 replicate populations each
with 8 boy 8 girl flies - all flies were
initially heterozygotes for a brown eye color
allele (bw/bw-75) - each generation, out of all
offspring, 16 were chosen to start the
next generation - monitored for 19 generations
38
Random fixation and loss of heterozygosity
Expected result - no selective advantage, so
bw-75 allele should drift to fixation 50
of the time and be lost 50 of the time -
heterozygosity should decrease over
time   Results after 19 generations - in 30
populations, bw-75 allele was lost - in 28, it
was fixed
39
Random fixation and loss of heterozygosity
Heterozygosity could also be directly scored by
eye color
- decreased every generation, as predicted by
theory - actually decreased faster than
expected, as though N 9 flies (not 16)
40
Founder Effect
When a population is founded by a few initial
colonizers, their genetic make-up will
largely determine the allele frequencies as
the young population grows A small group of
founders will not carry all the alleles present
in the larger population they came from
(reduced genetic diversity) - if founders carry
rare alleles, these alleles will be over-
represented in the new population relative to the
original large population   example
Pennsylvania Amish carry allele for a rare form
of dwarfism, at a frequency of 7 - only
present at 0.1 in most populations - one of
original 200 founders had the recessive allele
- consequence way more Amish dwarves than
youd expect
41
Genetic drift and Elephant tusks
98 of 174 female elephants in the Addo National
Park lack tusks
- population was reduced to 11 individuals by
hunting, until protected in 1931 - at that
point, 50 of females lacked tusks - near
loss of female tusks is likely a result of
genetic drift, following population
bottleneck
proportion of females w/ tusks
population size
42
Genetic drift and Elephant tusks
98 of 174 female elephants in the Addo National
Park lack tusks
- Alternative hypothesis ivory hunters
imposed strong selection against tusks - if
tusklessness were a recessive trait, what
would you expect to happen to the frequency
of tusklessness since the population was
protected? Why?
proportion of females w/ tusks
population size