Transcript Mechanisms of Evolution

Mechanisms of Microevolution
Reading:Freeman, Chapter 24, 25
Microevolution vs. Macroevolution
 The term “microevolution” applies to evolutionary change within a lineage
– It occurs continuously.
– Depending upon the organism and the circumstances, it can transform a lineage.
dramatically over time.
– Alternately, a lineage may appear to remain the same over time-this is called
stasis.
 Macroevolution is the origin and extinction of lineages.
– It can happen gradually, or slowly.
 Both processes are essential to evolution. Microevolution is probably better
understood, and better documented, because in some organisms it takes place
on timescales we can study directly by experiment and direct observation.
 Ironically, in “On the Origin of Species” Darwin lays out a theory of
microevolution…he assumed macroevolution would inevitably result from
microevolution.
– It would be 100 years later that Ernst Mayr, and others, would develop a scientific
theory of speciation.
– The replacement of one species by another (as opposed to the replacement of one
allele by another), by the way, is an ecological process..it is not evolution in the
usual sense, though this phenomenon usually leads to extinction of some species
and diversification of others.
The Population is the Basic Unit
of Microevolutionary Change
 The genotype of an individual is, essentially, fixed at birth.
 The population is the smallest unit where evolutionary change is
possible.
– Unlike individuals, populations permit the origin of new alleles
through mutation, and the change in the frequency of alleles
through selection, genetic drift, etc..
 Individuals do not evolve, populations and species evolve.
Population Genetics
 Population genetics refers to the study of evolution via
the observation and modeling of allele frequencies and
genetic change in populations of organisms.
 There are three parameters to keep in mind:
– allele frequency: the proportion of a specific allele at a given
locus, considering that the population may contain from one
to many alleles at that locus.
– genotype frequency: the proportion of a specific genotype at
a given locus, considering that many different genotypes may
be possible.
– phenotype frequency: the proportion of individuals in a
population that exhibit a given phenotype.
 Consider a population of N organisms.
 Two phenotpyes, yellow and tan.
 Suppose that they are diploid and reproduce sexually.
 Consider one gene with two alleles, A and a.
 The possible genotypes are therefore:
 AA, Aa, and aa.
Phenotype Frequencies
 To calculate the frequency of a phenotype, count the
number of individuals with that phenotype, and divide by
the total. Therefore, the frequency of the yellow
phenotype in the population below is 4/10=.40
Genotype Frequencies To calculate the frequency of a genotype in the
population, find the total number of individuals
in the population with that genotype, and
divide by the population size, N.
– f(AA)= #(AA)/N
– f(Aa)= #(Aa)/N
– f(aa)= #(aa)/N
 Question: What are the frequencies of the AA and Aa,
and aa genotypes in the population below?
Aa
Aa
aa
aa
AA
Aa
aa
aa
AA
Aa
 Answer: freq(AA)=2/10=.20 freq(Aa)=4/10=.40
freq(aa)=4/10=.40
Aa
Aa
aa
aa
AA
Aa
aa
aa
AA
Aa
 Allele Frequencies– By convention, the frequency of the dominant allele is
called p, thus the frequency of the recessive allele, q=1p.
 To calculate the frequency of an allele in the population,
add the total number of homozygotes for that allele to half
the heterozygotes, and divide by the population size, N.
– p= ((#AA) + (1/2)(#Aa))/N
– q= ((#aa) + (1/2)(#Aa))/N
 If you already know the genotype frequencies,
– p=f(AA)+(1/2)f(Aa)
– q=f(aa)+(1/2)f(Aa)
 Question: What are the frequencies of the A and a
alleles in the population below?
Aa
Aa
aa
aa
AA
Aa
aa
aa
AA
Aa
 Answer: freq(A)=p=(4+(2x2))/20=.40

freq(a)=q=(4)+(4x2)/20=.60 or q=1-.4
 since p+q=1
Aa
Aa
aa
aa
AA
Aa
aa
aa
AA
Aa
Evolutionary Change is a Consequence
of Changes in Allele Frequencies
 This is the genetic definition of evolution…a
synthesis between Mendel’s and Darwin’s
theories.
 All of the evolutionary change between our
single-celled ancestors and ourselves can be
described as the sequential origin of new alleles,
their replacement of old ones, and occasionally the
origin of new genes through duplication..
Evolution as Change in Allele/Genotype
Frequency
 Microevolutionary change is inherent in the change in
the frequencies of alleles over time.
 To be able to see if evolution IS occurring, we need to
consider what we would expect if evolution is NOT
occurring.
 Once we know that, then if we see a departure, we
know that evolution is occurring.
The Hardy-Weinberg Equilibrium
 Hardy-Weinberg Equilibrium is defined as the
situation in which no evolution is occurring. It is
a genetic equilibrium.
 It was the solution to a nineteenth century
misconception-the notion that the dominance or
recessiveness of an allele alone could cause
evolutionary change (it can’t).
 The Hardy-Weinberg Equilibrium refers to a
particular locus: one locus may be undergoing
rapid allele-frequency change, while other loci are
in equilibrium.
Assumptions of the Hardy-Weinberg Equillibrium
 A locus with two or more alleles will be in Hardy-Weinberg
Equilibrium if five assumptions are met. These are:
– 1. infinite population size (there are infinitely many individuals
in the population.)
– 2.there is no allele flow (i.e., no movement of individuals from
population to population.)
– 3.there is no mutation (no biochemical changes in DNA that
produce new alleles.)
– 4. there is random mating (this means that with regard to the trait
we're looking at, individuals mate at random they don't select
mates based on this trait in any way.)
– 5. No Selection: the different genotypes (for the genetic trait we're
studying) have equal fitness.
 Consider a population of diploid, sexually reproducing
individuals. Imagine a gene with two alleles, A and a, so
there are three genotypes, AA, Aa, and aa.
– Assume this population meets the five assumptions of HardyWeinberg Equilibrium.
– Because mating is random, alleles mix together at random.
– Because the population is infinitely large, the probability of
getting a gamete with a particular allele in it is simply the
frequency of that allele.
– Similarly, determining the probabilities of getting particular
genotypes will tell us the frequencies of those genotypes in the
population.
– To get AA, you need an egg with allele A, with
probability p, and a sperm with allele A, also with
probability p.
– The probability of getting both of these is p2
– Similarly, for aa, the chance of a sperm with an "a"
allele is q and the chance of an egg with an "a"
allele is also q, so: q2
 There are two ways of getting the heterozygous genotype,
Aa. These are:
 1.an "A" bearing sperm and an "a" bearing egg

2.an "A" bearing egg and an "a" bearing sperm
 The probability of an "A" bearing sperm is p; the
probability of an "a" bearing egg is q, so the probability of
the first way of getting an Aa is (p)(q) = pq.
 Similarly, the probability of aA (a sperm, A egg) is also
(p)(q)=pq.
 So to get the probability of getting the Aa genotype, we
add together the probabilities of the two ways of getting
this genotype. So: The genotype frequency of Aa is
pq+pq=2pq
– With a little more math, it can be demonstrated that:
As long as the conditions of the Hardy-Weinberg
equilibrium are met, allele frequencies will remain
constant.
• After one round of random mating, there will exist a stable,
mathematical set of genotype frequencies for any given allele
frequencies.
– This relationship will hold as long as the conditions of
the Hardy-Weinberg equilibrium are met.
• This means that if you know allele frequencies, you can solve
for expected genotype frequencies.
• If you know the frequency of a recessive genotype, you can
usually infer the expected frequency of the recessive allele as
well.
These apply under Hardy-Weinberg equilibrium:
Freq(AA)=p2
Freq(Aa)=2pq
Freq(aa)=q2
p2+2pq+q2=1
 So-Without the HW equilibrium, you can do thus:
 Allele frequencies from numbers of individuals
– p= ((#AA) + (1/2)(#Aa))/N
– q= ((#aa) + (1/2)(#Aa))/N
 Genotype frequencies to allele frequencies
– p=f(AA)+(1/2)f(Aa)
– q=f(aa)+(1/2)f(Aa)
 To go from allele frequency to genotype frequency, you must assume
HW equilibrium
– If this is not the case, there may be many potential genotype frequencies
for a single allele frequency.
– For instance, 500AA 0Aa 500aa, and 1000Aa both have p=.5, q=.5
 If HW is the case, use the following to go from genotype frequency to
allele frequency, but remember these are estimates (and expected
values of each would be these estimates times the total population
size), not the true values:
– Freq(AA)=p2
Freq(Aa)=2pq
Freq(aa)=q2
 If you have the frequency of a recessive phenotype, its frequency is q2, thus,
take the square root and get q. Get p from there, etc.
Question:
– The albino color in rabbits is caused by a recessive
allele. Aa and AA individuals are normally pigmented,
and aa individuals are albino.
– Imagine a population with 9999 normally pigmented
rabbits, and a single albino rabbit.
– What is the frequency of the recessive phenotype?
What is the frequency of the (aa) genotype?
– Under Hardy-Weinberg equilibrium, what are the
expected frequencies of the dominant and recessive
alleles? What is the expected frequency of the (Aa)
genotype?
Answer.
 Freq Recessive phenotype 1/10000=.0001
 Freq (aa) genotype, 1/10000, .0001
 q2=.0001, therefore q=(.0001)1/2 q=.01
 p+q=1, so p=1-.01=.99
 2pq=(.01x.99x2)=.0198
 Note that while 1 in 10,000 rabbits is an
albino, approximately one in fifty
individuals carry the allele. This is a fairly
common situation for a recessive allele
which is selected against.
 We can define five different ways in which evolution
occurs based on the situations in which these five
assumptions are NOT met. These are:
 1.genetic drift is change in allele frequency by
random chance. It occurs if a population is not infinite in
size. In populations that are not infinitely large, there
will be random error in which alleles are passed from
generation, and allele frequencies will change at
random. Since no population is really infinitely large,
there is always some genetic drift occurring; however,
the effect is very small in large populations. The effect
of genetic drift is larger in small populations.
 2. Allele flow is change in allele frequency that occurs
because individuals move among populations. If there are
different allele frequencies in different populations of a
species, then when individuals move to a new population,
they will change the allele frequencies in the new
population.
 3. mutation is biochemical change in DNA that one allele
into another and creates alleles. It not a common event
(typical mutation rates are about one mutation in a million
genes passed from generation to generation ); as a result,
evolution through mutation is extremely slow.
 Mutation is very important for evolution because,
ultimately, mutation is the source of genetic variation.
Other forms of evolution on cannot occur without genetic
variation.
 With regard to the fitness of alleles, mutation is random -- it
may produce alleles that result in high fitness (rare) or low
fitness (much more common), and the probability of a
mutation is independent of an evolutionary “need” for the
mutation.
 4.non-random mating is evolution that occurs
because individuals select mates based on their
characteristics.
 5.natural selection is evolution that occurs because
different genotypes have different fitness. More
about this later.
You can test to see whether a population is in
Hardy-Weinberg equilibrium, if you have
the numbers of individuals of each
genotype.
-The Chi-Square test is ideal for this:
-Generate expected values from HW expectations and
compare them to observed values.
-The “null hypothesis” in this case is that the population IS
in HW equilibrium, and thus, no evolution is ocuring at
that locus at the moment.
-A violation might signal that some force of evolution is at
work.
 For instance, imagine that you had a population of
river grapes Vitis riparia.
– There is an enzyme locus you find interesting,(could be
anything, superoxide dismutase for instance).
– It has two codominant alleles, SODF, for fast (because
DNA fragments of it migrate rapidly on a gel), and SODS
for slow, because DNA fragments migrate less quickly
on a gel.
– You don’t actually know much about evolution in this
species, and you want to know if this locus is evolving.
– You census 1000 individuals (at least you think so, they
are grapes).
– You get:
• 460 SODF SODF
• 23 SODF SODS
• 517 SODS SODS
– Is the population in HW equilibrium?
– So, there is no dominant locus, but we will call the frequency of
SODF p.
– p=(460+11.5)/1000=.47, thus q=.53
– Observed
Expected
• 460 SODF SODF
• 23 SODF SODS
• 517 SODS SODS
221 SODF SODF
498 SODF SODS
281 SODF SODS
– Thus Chisquare=approximately 910
• There is only one degree of freedom here (crazy, I know it seems, but
we generated three expected values from just p and q, so df=1)
– This is much lager than the critical value, so the locus is not in HW
equilibrium.
– Why?
– A statistical test cannot answer that…but there are not enough
heterozygotes. This (as you will learn later) is the signature of
inbreeding, and grape plants self-fertilize.
– Alternately, since grapes reproduce asexually as well as sexually,
this could be the differential success of two genotypes, both
homozygotes, at reproducing asexually.
The Mechanisms of Evolution
 Mutation
 Allele (gene) flow
 Selection
 Genetic Drift
 Nonrandom mating
– each one is, in essence, the result of a violation
of one or more of the assumptions of the
Hardy-Weinberg equilibrium
Mutation
 A mutation is a change in the organism’s DNA.
– Mutations may affect somatic (nonreproductive
tissue), or they may affect the germ line
(reproductive tissue). Except in clonal
organisms, somatic mutations cannot generally
be passed on.
– Evolutionary biologists are interested in
heritable mutations, the kind that can be passed
on to the next generation.
 A heritable mutation changes one allele into
another, sometimes creating an allele that is not
already present in the population.
 Some mutations create dominant alleles, some
create recessive or codominant alleles.
 Some mutations are harmful or lethal, many are
totally neutral-they have no effect, a few are
favorable.
 Whether a mutation is harmful, neutral, or
favorable, depends upon its environment
Some
types
of
mutations.
 Substitution: one nucleotide is substituted for another,
frequently this causes no change in the resulting
organism, sometimes the change can be dramatic.
 Insertion: DNA is inserted into a gene, either one
nucleotide or many. Sometimes, entire genes are
inserted by viruses and transposable elements.
 Deletion: DNA bases are removed.
 Small insertions and deletions can inactivate large
stretches of a gene, by causing a frame shift that
renders a gene meaningless.
 Duplication: an entire gene is duplicated.
 Transposition: DNA is moved to a new place in the
genome, frequently this happens because of errors in
meiosis or transposable elements.
Cat Mutations
Songbird mutations
 Mutations are random events: their
occurrence is independent of their selective
value - i.e., they do not occur when they are
needed any more often than they would
otherwise.
 Mutations at any single locus are rare
events: mutation rates at a typical locus are
about 1 in 106 gametes.
 Since each individual has thousands of
alleles, the cumulative effect of mutations is
considerable:
– Consider that each of us has about 3.5x104 loci,
and the mutation rates are about 1x10-6 per
locus, thus, about 1 in 30 of our gametes has a
new mutation somewhere in its genome. That
means about 7% of us are mutants, more or
less. YOU could be a mutant.
Mutations are the ultimate source
of genetic variation
 Mutations are the only source of new alleles
(other than the occasional transfer of alleles
by viruses).
 Mutation is thus the ultimate source of
genetic variation…it creates the raw
material upon which natural selection acts.
Example-an interesting mutation:
 In humans, one interesting mutation is called the CCR-
d32 allele (the locus is named CCR, it is one of many
alleles at that locus)
– This allele codes for a 32 base pair deletion that makes the
protein nonfunctional.
– Lacking this protein on the surface of their blood cells,
homozygous individuals (it is effectively codominant) are
essentially resistant to HIV-HIV cannot infect their cells.
 This mutation did not arise because of HIV, best we
can figure, it predates the evolution of HIV by
hundreds or thousands of years, and was neutral (or
possibly maintained by selection induced by the
bubonic plague) until HIV entered out species!
Genetic Drift
 Genetic drift is the change in allele
frequencies that occurs by chance events. In
essence, it is identical to the statistical
phenomenon of sampling error on an
evolutionary scale.
 It is a random (stochastic) process.
 Because sampling error is greatest in small
samples and smallest in large samples, the
strength of genetic drift increases as
populations get smaller.
Effects of Genetic Drift
 Does not generally result in adaptation.
 In a large population - it has little effect unless
enormous spans of time are involved.
(if vast spans of time ARE involved, the
cumulative effects of drift on any species can
be considerable-genetic drift is the primary
mechanism for the substitution of neutral
alleles over time, which is the mechanism of
the molecular clock used in systematics.)
 In a small population
 alleles can be lost (usually the
rare ones)
 other alleles are fixed-their
frequency reaches 1.0
 genetic variation is lost, resulting
in at population can become
homozygous at many loci
Founder Effect
The founder effect is genetic drift that occurs
when when a few individuals, representing a
fraction of the original allele pool, invade a new
area and establish a new population.
 Examples:
 California Cypress-a very large
population was established from a small
number of individuals.
 Founder effect occurs in many
introduced species.
 Amish-a religious minority, which is
essentially an isolated population,
established from a relatively small
number of individuals.
Bottlenecks
 Bottlenecks are periods of very low population
size or near extinction. This is another special
case of genetic drift.
 The result of a population bottleneck is that even if
the population regains its original numbers, genetic
variation is drastically reduced
 Examples:
 Cheetahs -nobody knows exactly why it
occurred, but cheetahs underwent an
extreme population bottleneck several
thousand years ago. As a result, they have
very little genetic variation.
 Northern Elephant Seal-underwent an
extreme population bottleneck resulting
from fur hunting in the nineteeth century.
 Ashkenazic Jews-a religious minority in
Central Europe that has rebounded from
attempted genocide.
 Endangered Species
 Genetic drift contributes to evolution in a number of
ways
 by decreasing genetic diversity, it can put the
population at risk of extinction
 its random nature increases the genetic
differentiation between two or more populations
 this may lead to speciation if one or more
populations become reproductively isolated.
 Genetic differentiation caused by genetic
drift may change the genetic background
against which new mutations act. If there is
epistasis, a new mutation may be favorable in
some populations and unfavorable in others.
(Wright’s shifting balance theory)
The Neutral Theory of Molecular Evolution.
 One of the most interesting breakthroughs in evolutionary biology in





the 1960’s-1990’s, has been the development of the neutral theory of
molecular evolution.
It was introduced by the Japanese theoretician Motoo Kimura, in the
late 1960’s.
It is a theory of evolution that Darwin never could have anticipated
(evolutionary biology does not begin and end with Darwin).
It runs in parallel with Darwinian evolution by natural selection,
though its effects are most noticeable and easiest to understand on loci
for which there are no differences in fitness between alleles (thus, it is
called the neutral theory).
It causes change over vast spans of time, at a more-or-less constant
rate, when averaged over many loci.
For that reason, it can be used to develop a “molecular clock”..to tell
how long it has been since two lineages have diverged.
 In the 1960’s techniques of observing genetic variation in natural
populations became available, and were pioneered by researchers such
as Richard Lewontin.
– It was discovered that, in natural populations, many selectively neutral
genetic polymorphisms exist.
 Kimura based his theory upon this.
– Thus, he hypothesized that much of genetic variation is actually neutral
– He also asserted that most evolutionary change is the result of genetic drift
acting on neutral alleles.
– New alleles originate through the spontaneous mutatation of a single
nucleotide within the sequence of a gene.
• In single-celled organisms, or asexuals, this immediately contributes a new
allele to the population, and this allele is subject to drift.
• In sexually reproducing, multicellular organisms, the nucleotide substitution
must arise within the germ line that gives rise to gametes.
– Most new alleles are lost due to genetic drift, but occasionally one
becomes more common, and by random accident, replaces the original.
– The chance of this is small, but over time, it happens occasionally, at a
predictable rate.
 In this way, neutral substitutions tend to accumulate, and genomes tend
to evolve.
– Many of the polymorphisms we see may be “transient”-one allele is in the
process of replacing another.

*Stolen from a great site nitro.biosci.arizona.edu/.../Lecture47.html
 In the scematic diagram above, you can see that some alleles are lost
over time, but occasionally, one becomes fixed, and replaces the other,
all by random genetic drift.

*Stolen from a great site nitro.biosci.arizona.edu/.../Lecture47.html
 Although its importance, relative to Darwininan evolution, is debated, this theory is
farily well supported by now.
 Rates of molecular evolution vary among proteins, and among organisms. Some
proteins allow much less neutral variation, and evolve more slowly.
 Interestingly, population size is not that important for rates of molecular evolution (it
cancels out in the math, small populations drift faster, but have fewer mutants per
generation)
Population Structure
 Most species do not exist as many
populations, which are isolated from each
other to some extent.
– Populations occasionally exchange members.
– Most populations are spatially structured;
individuals tend to cluster in areas of suitable
habitat.
• These local aggregations, called subpopulations,
regularly exchange members.
Allele Flow:
Allele flow (or gene flow) is an evolutionary
force that results from migration of
individuals or the dispersal of seeds, spores,
etc.
Allele flow can potentially cause evolutionary
change, provided that:
1) the species has multiple
subpopulations.
2) there are differences in allele
frequency among populations, or among
subpopulations within populations.
Effects of Allele Flow
 Even small amounts of allele flow can negate
genetic drift.
 If natural selection favors certain alleles in
some populations, and different alleles in
others, allele flow can oppose natural
selection and prevent the evolution of genetic
forms suited to each environment.
 If sufficiently strong allele flow will cause
allele frequencies in different populations to
converge on a single, population-wide mean.
Allele Flow v.s. Genetic Drift
 When does allele flow negate the effects of
genetic drift?
 Let us consider a whole bunch of semi-isolated
populations, that exchange occasional migrants
with each other; exchange of migrants is random
 Let m=the proportion of migrants exchanged per
generation.
 Let N=the number of individuals in each
population.
 I will spare you the proof, although famous,
it runs several pages….
 Allele flow will negate the effects of genetic
drift if m>(1/(2N)).
 This is a very small number, one migrant
every other generation is sufficient to
prevent genetic drift from causing
evolutionary differences among populations
of a species, or subpopulations within a
populations of a species.
Allele flow and selection
 Note that allele flow can also oppose selection.
 On the edge of the range of a species, there might
be local populations adapted to special conditions.
 Allele flow from a large, central population
adapted to a different environment might swamp
the effects of natural selection, by causing an
influx of less fit alleles every generation to
counterbalance the unfit alleles lost to selection.
Nonrandom Mating
 Two important patterns of nonrandom
mating affect evolution:
 1) Inbreeding, or mating between relatives
(selfing is a form of inbreeding)
 2) Assortative Mating
Inbreeding
 Inbreeding, including selfing, is common in many
species. Inbreeding was formerly common in
humans, before the advent of increasingly
sophisticated forms of transportation.
 High levels of inbreeding lead to the loss of the
heterozygous genotype, although allele frequencies
are not necessarily changed.
 Inbreeding exposes recessive alleles to selection,
since they are more likely to be present in the
homozygous state if the population is inbred.
 Inbreeding can cause a dramatic decline in
the fitness of a population, possibly
extinction, because many species harbor
numerous deleterious recessive alleles that
are effectively hidden from selection (i.e.
the Florida Panther), although other species
are unaffected by inbreeding (i.e., certain
groups of parasitic Hymenopetera).
Assortative Mating
 Assortative mating occurs when individuals
choose their mates based on their
resemblance to each other at a certain locus
or a certain phenotype.
 Positive assortative mating occurs when like
genotypes or phenotypes mate more often
than would be expected by chance.
 Negative assortative mating occurs when
similar genotypes or phenotypes mate less
often than would be expected by chance.
Examples of assortative mating
in humans
 Dwarfs: very high positive assortative
mating, individuals with achronoplastic
dwarfism pair up much more often than
would be expected by chance
 IQ: slight positive assortative mating
 Height: slight positive assortative mating
 Redheads: moderate negative assortative
mating-red haired individuals pair up less
often than would be expected by chance.
Natural Selection
 What is it? Natural selection is the differential
survival and reproduction of individuals with
certain traits.
 It acts on phenotypes.
 Because most phenotypes are, in part,
determined by an organism’s genotype at one or
several loci, natural selection has the potential to
cause change in the frequency of alleles through
time.
 Any allele that affects the ability of an
organism to survive and reproduce will be
subject to natural selection.
 In populations, natural selection operates
whenever individuals in the population vary
in their ability to survive and reproduce.
 Natural selection causes evolutionary
change whenever there is genetic variation
for traits that affect fitness.
 For Natural Selection to Operate:
 1) there must be variation
 2) some of the variation must affect survival
and reproduction of individuals
 For Natural Selection to Cause
Evolutionary Change
 1) there must also be allelic variation for
characteristics that affect fitness.
What is Fitness?
 Fitness is the ability of an individual to
survive and make copies of its alleles that are
represented in the next generation.
– The fitness of an individual organisms is
essentially the same as its lifetime reproductive
success. The fitness of a genotype is the average
fitness of all the individuals in the population that
have that genotype.
 It is NOT physical performance.
 Differences in fitness may be due to
differences in survivorship, differences in
fecundity, or both.
Absolute fitness vs. relative
fitness.
 An organism’s absolute fitness is the total
number of surviving offspring that an
individual produces during its lifetime (its
lifetime reproductive success).
– some things that contribute to fitness are: its
chance of living to a certain age (age specific
survivorship), its number of offspring during a
certain time interval (age specific fecundity).
– These are called components of fitness.
Relative Fitness
 For mathematical purposes, absolute fitness
is standardized to get relative fitness.
 Imagine there are several genotypes, each
codes for a different phenotype. The
genotype with the highest absolute fitness
has a relative fitness of 1.0
 For every other genotype, their relative
fitness is: absolute fitness of that
genotype/absolute fitness of fittest genotype
Example
 A beetle is polymorphic for color, it comes in black,
brown and yellow color morphs. Birds and lizards
prey upon them, so that, due to differences in
survivorship, the fitnesses of the color morphs differ.
 CBCB=black
67% chance of survival to adulthood
 CBCY=brown 93% chance of survival to adulthood
 CYCY=yellow 11% chance of survival to adulthood
 QUESTION: Assuming they have identical numbers
of offspring, what is the relative fitness of each
genotype?
Answer
 The relative fitness of brown is the highest, so
w(CBCY)=1
 the fitnesses of black and yellow are:
 .67/.93=w(CBCB)=.72
 .11/.93=w(CYCY)=.12
– Note how we designate fitness as w, to avoid
confusing it with frequency.
– Also note that survivorship is a component of fitness, it is
necessary to assume fecundity is identical in the three
morphs to calculate relative fitness.
– s= the selection coefficient, is the difference in
the fitness of a genotype between its own value,
and the ideal of 1.0.
– So, for a deleterious, recessive allele.
– w (AA)=1.0, w (Aa)=1.0, and w(aa)=1-s
Directional Selection
 Most extreme phenotype is the most fit.
When applied to a single locus, that means
one allele becomes more common until it
reaches fixation (I.e., frequency 1.0)
– Directional selection tends to eliminate genetic
variation over time. If directional selection is to
proceed for a long time, new mutations must
replace lost genetic variation.
– In laboratory experiments, directional selection
causes rapid change in phenotypes, followed by a
plateau, caused by the loss of genetic variation.
Directional
Selection
Example of Directional Selection: The
peppered moth, Biston bettularia
 One of the best known cases of directional
selection is the evolution of “industrial melanism”
in this species.
 It has two forms: dark (melanic) and light.
 Controlled by a single locus with two alleles.
 Melanic is dominant, so that MM and Mm are
dark, mm is light
 the moth rests on trees during the day, and uses
crypsis as protection from predation by birds.
 Kettlewell (1955) showed that the two forms
differ in their suceptibility to bird predation.
 The melanic form was rare in 1848. When it was
first reported outside Manchester, it was visible
against the lichen-covered trees and often eaten by
birds.
 Museum collections indicate that by 1898, the
melanic form had increased from <1% to >98% of
the population
 Soot had darkened the trees, making the light form
most visible.
 In rural areas with no soot, the melanic form was
still rare.
 (Since the passage of clean air laws in Britain, the
trend has reversed, and the light form is more
common once again.)
 Many laboratory experiments have documented evolutionary
change by imposing directional population on strains of
laboratory organisms.
– For instance, by imposing directional selection on Drosophila
melanogaster, researchers, such as Michael Rose and Brian Charlesworth,
have documented changes in body size, sternopleural bristle number, and
life history characteristics such as age at reproduction.
– Selection in laboratory populations often produces dramatic change,
quickly, followed by a plateau, as genetic variation is exhausted.
www.biology.duke.edu/rausher/lec11.htm
Stabilizing Selection
 Intermediate phenotypes, somewhere close
to the mean are most fit. When applied to a
single locus, it implies that the heterozygous
genotype is most fit, and is called Balancing
Selection.
– Generally, stabilizing selection maintains the
mean value for the trait, and decreases the
variation for the trait (thus, it usually decreases
genetic variation). In the special case of
balancing for a single locus, genetic variation is
actually preserved, since both alleles will be
maintained.
Stabilizing
Selection
Examples of Stabilizing Selection
 Stabilizing selection is probably common in
nature.
 Birth Weight in Humans: It is well known that
early mortality is highest for extreme birth
weights. Both very small and very large infants
suffer high mortality.
 Clutch Size in Birds and Parasitoids: Females
that lay intermediate numbers of eggs have the
highest reproductive success. Too many, and the
offspring all starve. Too few, and the mother
could have laid more. Called the Lack optimum,
it applies to many birds, also to parasitoids.
Example of Balancing Selection, and of the
Differing fitness of an Allele in Different
Environments: Sickle Cell Anemia in Humans
 Sickle cell anemia in humans is caused by an
allele that causes hemoglobin to deform under low
oxygen conditions, causing the red blood cell to
“sickle”.
 Homozygotes for normal hemoglobin Hb+ Hb+
have no illness.
 Homozygotes for the sickle allele HbSHBShave a
very serious genetic disease.
 Heterozygotes HbS Hb+ appear normal, but
occasionally their blood cells sickle under stress.
This is not particularly debilitating.
 In countries without malaria:
 There is strong selection against homozygotes for
the sickle cell disease, w(HbSHBS)=0, because, they
rarely survive long enough to have many offspring.
 The other two genotypes have a the same relative
fitness w(Hb+ Hb+)=w(HbS Hb+ )=1, because carriers
are essentially indistinguishable from those
possessing normal hemoglobin.
 There is thus directional selection against the sickle
cell allele.
 In countries WITH malaria:
 Heterozygotes for the sickle cell allele have some
limited resistance to malaria, because the cells
sickle and kill plasmodium within.
 The heterozygote is most fit w(HbS Hb+ )=1,
w(Hb+ Hb+)=.90, w(HbSHBS)=0.
 Selection acts to balance maintain both alleles
because the heterozygote is favored, this is an
example of balancing selection.
 This explains why the original distribution of the
sickle cell allele roughly matches the worldwide
prevalence of malaria.
Disruptive or Diversifying
Selection
 Two or more phenotypes are most fit, but
the intermediates have low fitness.
– Not particularly common in nature. In most
cases, it increases the variance for a trait, while
not affecting the mean. Combined with
assortative mating, it has the potential to form a
polymorphic population.
Frequency-Dependent Selection
 Selection can be frequency-dependent
 + frequency dependent selection: most
common type is the most fit
 - frequency dependent selection: least
common type is the most fit.
– Example-Eye-eating cichilids in Lake Victoria.
– Example-Color polymorphism in elderflower
orchids. Interesting case studied by Luc Gigold
et al.
 Elderflower orchids have two colors, yellow
and purple.
 Populations typically have both color morphs,
generally with the yellow morph being slightly
more common.
 Bumblebees are the primary pollinator.
– Like many orchids, elderflower orchids are
deceptive. They advertize to bees, but offer no
nectar reward.
– Bumblebees learn to recognize the most common
morph, and learn to avoid it, giving an advantage to
the least common morph.
– Gigold demonstrated this by experiment,
planting arrays of orchids where the color
morphs occurred in different frequencies.
– The least common morph had higher
reproductive success, whichever form that was.
– Thus, selection is -frequency dependent.
 Natural Selection has been documented and studied in nature
many times.
– Despite Darwin’s intuition, many documented examples of natural
selection in the natural world show rapid evolution and a dramatic
response to natural selection, rather than slow, gradual change.
– Interstingly, since the environment changes, selection in the real world
often reverses direction and is not consistent over time and from one
location to the next.
– The best know study of natural selection in the wild was the study of
Galapagos finches by Rosemary and Peter Grant, and their colleagues.
– They documented dramatic changes in finch populations, in response to
strong natural selection imposed by drought.
The Environment affects the
Fitness of Alleles
 Alleles may have different fitnesses in different
environments. An allele that is favored in one
environment may have a disadvantage in another
environment.
 For systems of epistasis, the genetic environment
may affect the fitness of an allele.
 The frequency of an allele may also affect its
fitness.
 Examples: Sickle cell anemia, the Lap locus in
mussels, color patterns in Heliconia butterflies,
chromosomal inversions in Drosophila.
 Environmental change may reverse the effects
of selection.
– Over evolutionary time, this seems to be the rule
rather than the exception.
 Selection has no memory, no plan, and no
goal.
– There was no special driving force in evolution to
produce human beings, or anything like us. This
does not exactly make us an “accident:”, more
precisely, it makes us one species among billions
of potential evolutionary outcomes.
 Selection does not act for the good of the
species, nor for the good of the planet.
Selection is weak against rare
recessive alleles
 As you can see from the preceding
equation, as recessive alleles become rare,
selection against them becomes weaker and
weaker, because most copies are likely to
exist in the heterozygous state
 Likewise, selection in favorable of new
mutations is very weak if that mutation is
recessive. Favorable, recessive mutations
can be lost by genetic drift before they have
a chance to spread by selection.
Mutation-Selection Balance
 Imagine that allele A mutates into
disadvantageous allele a at rate u.
 u is usually very small, on the order of 10-8.
 Selection will reduce the frequency of a to a
low level, but selection is weak against
uncommon recessive alleles.
 A mutation-selection balance will be reached
where
 q*=equilibrium frequency of allele a=(u/s)1/2
 This result has enormous significance to medicine. It
provides an explanation for the continued existence
of alleles which, when homozygous, cause severe
hereditary illnesses….most of these are in mutationselection balance.
 Alleles which occur in unexpectedly high
frequencies in some populations, such as sickle cell
anemia, thallassemia, or cystic fibrosis, may have
been subject to balancing selection in the past.
– The agent of selection for Sickle Cell Anemia, and for
thallasemia (Mediterranean populations of humans), was
probably malaria, for CF, it was most likely typhus.
 Alternately, it is possible that genetic drift has caused
them to become more common that would be
expected under this model.
– This may be the mechanism causing Tay Sachs disease to
be unexpectedly common in certain Jewish populations.
 Behavioral Ecology is the science of the
ultimate, evolutionary causes for behavior.
 It brings together three sciences; ecology,
animal behavior, and evolutionary biology.
– It is a very new science, resulting from the
intersection of ideas by several influential
schools of thought
– These include; David Lack, who pioneered the
comparative approach, W.D. Hamilton, who
pioneered the concept of kin selection, and E.O.
Wilson, who pioneered sociobiology.
 Behavioral Ecology has 2 basic themes:
 Natural selection maximizes the ability
of an organism to survive and
reproduce. Individuals (in essence,
temporary vehicles for genes and
alleles) should behave in ways that
maximize inclusive fitness.
 The so called “optimal” behavior
needed to maximize inclusive fitness
will depend on both the behavior of
other individuals & ecological
circumstances.
 Evolution should select for
behaviors that enhance fitness.
But, will behaviors always be
optimal?
 Possible 'obstacles' to optimal behavior & fitness:
 Mutation-many mutations produce individuals with lower fitness.
 Linkage - genes beneficial in some way may be very close on a
chromosome to a gene that tends to reduce fitness. So, to get the benefit
of one gene, an organism must withstand the liability of the other.
 Pleiotropy-Alleles have multiple effects. So, if an allele influences traits
X, Y, & Z, with X being an optimum phenotype, there's no reason to
assume Y & Z are also optimum
 Variable environments - difficulty of achieving optimal behavior varies
in proportion to variability of the environment
 Evolutionary lag-individuals adapted to past conditions are not
necessarily adapted to present conditions
 Phylogenetic inertia-evolutionary baggage'; resistance to acquisition of
adaptive characteristics due to prior evolutionary history (e.g., flightless
birds on many islands)
The Comparative Method
 The comparative method: comparing species
in divergent lineages, to see if there is a
pattern of convergent evolution, where
lineages that enter particular ecological niches
evolve certain behaviors or structures, is
widely used in evolutionary biology.
 By comparing different species, behavioral
ecologists can link behavior and social
organization to ecological factors.
The Comparitive Approach is
Good for Making Inferences
About Evolutionary Trends.
 For example, why is
male mortality greater
than female mortality
in some species?
 For instance, once
reproductive maturity
has been reached,
male mortality is much
greater than female
mortality in orcas.
 This pattern is present among humans as
well, but is very slight.
 John Allman hypothesized that it was related to parental care.
Natural selection favors increased lifespans for males in species
where males provide parental care.
Criticisms of Optimality Models
 The major criticism of optimality models is that they appear
to test the animal to see whether it is behaving perfectly
according to the model.
– It is important not to loose sight of the fact that the model
is being tested, not the animal.
 The inherent assumption of optimality theory, that animals
behave in a way that maximizes their fitness, is not likely to
be met all the time.
– Thus, if the animal’s behavior does not match the model,
is the animal flawed or is it the model?
– One of the best approaches is to see whether animals that
behave according to the model have higher fitness than
animals that deviate from it.
Ecological Game Theory
– One of the most interesting things about natural
selection for a behavior is that the behaviors of other
individuals in the populations may affect the fitness of
an individual.
– Some behaviors, such as stealing from another
individual’s nest, may be favored by selection when
they are uncommon (thieves are rare and the population
is naïve). When they become common, however, they
may be selected against (in a population where
everyone is stealing from everyone else, there is
nothing left to steal)
– Axelrod, Hamilton, John Maynard Smith, and others
brought the economic science of game theory over into
behavioral ecology to explain how these systems work.
Example of an Ecological “Game”, Stealing
in Digger Wasps
 The great golden
digger wasp
(Sphex
ichneumoneas) is a
close relative of
the species Fabre
studied.
 It digs a burrow in
the ground and
stocks it with
paralyzed crickets.
 When provisioning nests, wasps can pursue
two alternative strategies.
– Dig a burrow.
– Enter pre-existing burrows and take them over.
– (A “strategy” is a behavior whose fitness depends
upon the behaviors of others.)
 The second strategy has advantages when it
is rare.
– It is easier to take over a burrow, it takes less
time and involves less wear and tear to the
wasp.
 As this behavior becomes more common,
however, disadvantages accumulate.
– With more wasps looking for pre-existing
burrows, there are no “empty” ones left over
when wasps die, and most of the time the
individual has to fight a wasp for the nest. This
is time-consuming and potentially deadly.
 Ecological game theory predicts that the two
behaviors should be “balanced” in the
population by natural selection, so that they
both have equal fitness.
– If one has a higher fitness, it will become
progressively more common until its fitness drops
enough to equal the other one.
– Wasp fitness is relatively easy to measure: by
counting cocoons (the larvae eat the cricket and
spin a cocoon), it is possible to assess how many
offspring they have produce
– Thus, by observing wasps in nature, it should be
possible to test this prediction.
 Brockmann and Grafen tested the prediction
by observing the wasp over the course of
several seasons.
– Their result; the two strategies have roughly
equal fitness (though a wasp will switch from
one strategy to the other-thus the behavior is
not strictly encoded by genes, but genes
presumably determine a flexible strategy where
individuals change tactics based upon their
environment to maximize their fitness)
 Altruism is behavior that benefits another
individual at the expense of one’s self.
– From an evolutionary standpoint, the trouble with
altruism is that if an altruistic behaviour is costly for
example, dying for someone else then the genes that
promote it should quickly disappear from the
population.
– Yet examples of self-sacrifice abound in animal
societies. The most conspicuously selfless are the
social insects, the ants, bees and wasps in which most
individuals work tirelessly for the good of the colony
and never reproduce themselves. How can such
behavior be explained?
– W.D. Hamilton argued that such extreme
altruism is most likely to evolve if, by
sacrificing themselves, individuals increase
their "inclusive fitness"that is, the proportion of
their genes carried by others in the population.
"Hamilton's rule" is the mathematical
formulation of this; put crudely, it amounts to
the idea that you can die for close kin and still
spread your genes, since close kin have many
genes in common with you.
Kin Selection
 Some behaviors have evolved to increase
the fitness of an organism’s close relatives.
 Individuals with certain heritable traits
might have genotypes that code for
behaviors that help close relatives raise
more offspring than they would without
help.
 These alleles would be favored by selection,
provided that their close relatives are likely
to have copies of the same heritable traits.
 The total copies of the gene coding for the trait
increase, even if the allele does not confer an
increase in fitness to the individual in question.
 The potential for kin selection has been
documented in many organisms, it is best known
in ants, bees and wasps.
 Relatedness is the probability that a particular allele,
present in one individual, is also present in another
individula because of descent form a common
ancestor.
 r=the coefficient of relatedness, is a mathematical
expression of that probability.
 It is easy to calculate r, given that you know how two
individuals are related.
 EXAMPLES:
– Identical twins - r = 1
Unrelated individuals - r = 0
Parent & offspring - r = 0.5
Full siblings - r = 0.5
Half siblings - r = 0.25
Uncles/aunts & nieces/nephews - r = 0.25
Cousins - r = 0.125
Hamilton’s Rule
 A behavior that helps a close relative’s
fitness, even though it harms one’s own
fitness, should be expected to evolve
provided that:
Br>C
 Where:
– B is the benefit to the relative (in terms of
increased fitness)
– r=the coefficient of relatedness
– C= the cost of that behavior in terms of one’s
own fitness.
 An organism’s inclusive fitness is a function of
one's behavior on their own survival and
reproduction, as well as the effect of one's
behavior on all relatives (with the importance of
each relative valued in proportion to degree of
relatedness).
Example; multiple foundresses in Polistes.
 Polistes sp. are a genus
of paper wasps.
– Any female has the
potential to be a queen,
though those born at the
end of the year are larger
and more dominant.
– Small females born in the
middle of the summer are
always workers, though
one of them will
potentially take over if
the queen dies.
– These large females mate and overwinter. In spring,
they emerge and;
• found their own nests
• join females that have already founded nests.
– Females that “join” usually get bullied into the
subordinate, worker role.
• Thus, their personal fitness is close to zero (they occasionally
sneak a few of their own eggs in).
• Thus, C is the number of offspring their colony would have
produced if they did not join another female’s nest.
– In the worker role, they increase their own
inclusive fitness by helping the queen.
• Nests with multiple foundresses are more likely to
survive and produce more males and potential queens
at the end of the year (fitness for a colony of social
insects must be measured by counting males and
queens, since workers do not reproduce their genes).
• Thus, B is the increased number of offspring a colony
will have when it has the extra foundress.
 Kin selection theory predicts that wasps should
join if Br>C
– Thus, unrelated females (r=0) should not work
together because there is no benefit to the female who
does not become queen.
– Wasps have a special genetic system, so that sisters
are particularly closely related. r between wasp
sisters is between .75 and .50, depending upon
whether they share the same father.
 Various researchers have studied this system (and
others, including the related wasp
Mischocyttarus) to see if Hamilton’s rule is
supported.
 Studies of Polistes support Hamilton’s rule.
– Unrelated sisters do not generally cooperate (at least
not after the issue of who will become queen is settled)
– Lone females have a low chance of colony survival.
– The first female who joins her sisters and acts as a
“worker” definitely increases the chance of colony
survivorship (thus, B is greater than zero).
• It is tough to measure B exactly, and one of the reasons
females join is to possibly take over the nest from the female
who founded it, thus there is an element of game theory to
the system.
• Accounting for the fact that the female that looses the battle
would have very low fitness if she went off on her own,
because she would be a late-starter (thus, C is small), Br>C.
– Extra helpers after the first “joiner” help the colony
less
• (thus, B is low for the second joiner, lower still for the
third joiner).
– In Northern North America, you rarely see more
than two Polistes females working together.
– In environments where there is competition for nest
sites, or colony founding is particularly dangerous,
more queens will work together, because B is larger
(reflecting the extra need for help) and C (reflecting
the smaller sacrifice in fitness because going it alone
is so risky) is smaller.
• Thus, several Mischocyttarus (Caribbean species) females
work together, and hundreds of Polybia (closely related
genus, Central American) females may work together to
found a nest.
Three Mischocyttarus females
at a new nest, these are probably
original foundresses.
The first true “workers” are in
the cocoons.
This Polybia nest
was probably founded
by a dozen or more females
 How does an organism 'know' its relatives (and, if known or
recognized, the degree of relatedness)?
– In some cases, the biology of the organism facilitates
close relatives living in close proximity.
– In other cases, there are mechanisms of kin recognition.
– Kin recognition is sometimes accomplished by smell,
sometimes it is a function of cognition and memory,
sometimes it is biochemical.
• Social insects, such as Polistes, generally use smell. Colonies
have a distinctive “colony odor”. Thus, sisters can smell each
other.
Example-tunicates recognize each other by chemical
signals (MHC genes, the same chemicals that enable
our bodies to reject foreign organs) on their skin.
Kin colonies of tunicates grow together. Non-kin
colonies form a “zone of death” between each other.