1. Viability - the ability to produce adults from zygotes (survival).
2. Sexual Selection: The ability to obtain mates. Can result in sexual dimorphism.
3. Fecundity: the ability to produce healthy gametes.

There are three patterns of natural selection: stabilizing selection, directional selection, and diversifying selection.

A. Directional Selection

In directional selection, phenotypes at one extreme die off or fail to reproduce and those at the other extreme leave a higher number of offspring. This shifts the frequency distribution in the direction of the other extreme. Directional selection occurs when there is a change in the environment such that the phenotype at one extreme is not adapted to the new conditions while those at the other extreme have increasing survival and reproduction.

B. Stabilizing Selection

In stabilizing selection, individuals with extreme characteristics die off or fail to reproduce resulting in populations of individuals with intermediate characteristics.

Stabilizing selection is most common in unchanging environments.

C. Diversifying Selection

In diversifying (or disruptive) selection, extreme phenotypes become more frequent from generation to generation because they survive to reproduce while individuals with intermediate phenotypes die or fail to reproduce.

Diversifying selection promotes dimorphism (two forms of a trait) or sometimes polymorphism (many forms of a trait) in a population. This may happen in a diverse or cyclically changing habitat where it pays to have different forms adapted to different parts of the environment.

---

The Genetics of Natural Selection

Natural selection is the differential reproductive success of individuals with the phenotypic characteristics induced by the genotypes such as AA, Aa and aa. Simply put, if a mutant carrying aa could produce more offspring than individuals with the dominate phenotype, the a allele will increase in the population. In other words, when selection is favoring a genotype it will show an unexpectedly high frequency and, when it is acting against a genotype, it will show an unexpectedly low frequency.

There are basically 4 types of selection:

1. Selection against the dominate phenotype (AA or Aa)

2. Selection against the recessive phenotype (aa)

3. Selection against the homozygotes (AA or aa)

4. Selection against the heterozygotes (Aa)

1. Selection Against the Dominate Phenotype (AA and Aa genotypes)
An extreme case - the dominate allele is lethal (causes death). First consider what would happen if the death occurs before reproductive age: In this case the dominate allele will be removed from the population in one generation.
For example, if we had:
AA X Aa X aa

and A becomes lethal due to an abrupt environmental change, what would happen in the next generation?:

The only offspring to survive to grow up and produce offspring of their own are those with the genotype aa.

So in the next generation it will be:

aa X aa

All of the offspring, of course, will be aa and we say that the a allele is fixed in the population. Even if the environment changes back and A is no longer lethal, A will not reappear.

How many dominate alleles will be passed on if selection isn't totally against the dominate allele or if death does not always occur before reproduction?

At this point the concept of fitness and the selection coefficient need to be defined:

1. Fitness: (W) is the ratio of offspring that survive to reproduce to the offspring that we expect to survive if there is no selection.

For example:

Parents: AA Aa aa in a 1:2:1 ratio with p=q=0.5

Expected (if the population is at Hardy Weinberg equilibrium):
Offspring: AA Aa aa in a 1:2:1 ratio with p=q=0.5

If 200 offspring are produced, we expect when they reach maturity to have:

50AA : 100Aa : 50aa

but when the F1 come to reproduce, suppose we only observe:

25AA : 50Aa : 50aa
(Expect 150 with the dominate phenotype but only observe 75)
Fitness = W = 75/150 = 0.5

If an allele is perfectly fit, then W=1.0; if it is totally lethal W=0.0.

2. Selection Coefficient: (s) is equal to 1-W.

In the above example, s = 1 - 0.5 =0.5
(This means that 50% of the dominate alleles are removed by natural selection.)

To determine the exact effect on frequency, we use the formula:

Where:

small triangle means change in frequency p
s is the selection coefficient against the dominate allele,
p is the frequency of the dominate allele in the parent population, and q is the frequency of the recessive allele in the parent population.

In the above example:
s = 0.5
p = 0.5
q = 0.5

In the next generation:

Therefore, in one generation of selection against the dominate phenotype the allele frequency of A drops from 0.5 to 0.4. What happened to the frequency of the recessive allele?

2. Selection against the Recessive Allele

a. Recessive lethal alleles: A recessive allele will only be selected against when it is homozygous i.e. aa -- this is the only condition when the recessive phenotype would be expressed in the environment.

All the a alleles did not disappear and the A alleles did not become fixed in the population. For the next generation, the following crosses are possible:

To calculate the frequency of the recessive allele in the next generation when selection is removing it from the population, we use the following formula:

where:

triangle q is the change in the frequency of the recessive allele after one generation of selection on it.
s is the selection coefficient for the recessive allele

Example:

A population where p=q=0.5 in the parental population and the genotypes present in a 1:2:1 ratio produces 400 offspring. The offspring still alive one week after birth are:

100AA 200Aa 0aa
We expected the genotypes to be in a 1:2:1 ratio -- in other words we expected 100 aa offspring.


What has been the change in the frequency of the allele in this one generation??


Changes in q are greater (more rapid) when the frequency of a is high than when it is low. Why? because at high frequencies there are many aa individuals in the population against which selection can act. At lower frequencies, change is progressively slower because there are fewer aa individuals. So as selection acts on the recessive, it progressively has less and less effect on the population as a whole.

3. Selection Against the Homozygote

Selection against the homozygote (AA or aa) is also called overdominance. It only occurs when the alleles were codominate or there was incomplete dominance. Otherwise the heterozygote would not be distinguishable from the homozygous dominate.
Lets look at what would happen if there was selection against the AA and aa genotypes:

The only thing left is the genotype Aa and in the next generation, all the matings will be Aa X Aa. The resulting offspring will be 1AA : 2 Aa : 1aa
In contrast to what we saw in selection against the dominate and recessive traits, both alleles are preserved because the heterozygote has both.

The formula for calculating the frequency of alleles when there is selection against the homozygotes is:

In fact there will be an equilibrium point at which both alleles will be preserved at a constant frequency. This equilibrium will be reached when the frequencies no longer change. In other words, when q = 0.
When --



This can be reduced:



Lets look at an example:

If selection against A is 1 (sA = 1) and against a is 0.2 (sa = 0.2), what will the frequencies of the alleles be when they reach equilibrium?

Since q = 1-p, then q = 1 - 0.16 = 0.84 at equilibrium.

The result of selection against the homozygotes is called a balanced polymorphism. "Polymorphism" refers to the three phenotypes (e.g., AA, Aa, and aa) that are retained by selection for the heterozygote and "balanced" refers to the equilibrium -- the tendency to preserve a constant ratio among phenotypes. A classic example of balanced polymorphism in human populations involves the gene for sickle cell anemia:

4. Selection Against the Heterozygote

In this situation, as in the case of selection against the homozygotes, there must be incomplete dominance or codominance of the two alleles so that the heterozygote can be phenotypically distinguished from the homozygotes.

Lets look at what will happen if the heterozygous condition is lethal:

In the next generation:


This type of selection results in the allele that is lowest in frequency at the onset of selection being lost and the allele in highest frequency being fixed. Lets look at a contrived example and see if we can understand this process:

Population with 100 individuals
30 AA : 50 Aa : 20 aa

There are 110 dominate A alleles (30 + 30 + 50). There are 90 recessive a alleles (20 + 20 + 50).

The 50 Aa individuals die but the 30 AA and the 20 aa mate. The 60 A alleles and the 40 a alleles may combine so that:

In the resulting generation the number of recessive alleles has been halved! If these individuals mate:



If these individuals mate:



All the recessive alleles removed.

The formula for calculating the change in frequency in one generation when there is selection against the heterozygote is:

---

Summary of selection on allele frequencies

1. When selection acts against the dominate phenotype, the dominate allele will be quickly removed from the population. If the dominate phenotype is lethal before the individual reaches sexual maturity, the dominate allele will be removed in just one generation and the recessive allele will be fixed.

2. When selection acts against the recessive phenotype, the recessive allele will be removed from the population over time. Even if the allele is lethal, it may take several generations to remove it from the population since it will not be selected against in heterozygous individuals.

3. In selection against the homozygotes, an equilibrium between the alleles is reached.

4. In selection against the heterozygote, the allele in lowest frequency at the onset of selection will eventually be removed from the population.

---

Fisher's Fundamental Theorem
Sir Ronald Fisher, a British statistician and pioneer in population genetics, showed mathematically that the greater the genetic variability upon which natural selection can act, the greater the expected improved fitness. Another way to view this idea is the rate of evolution can be limited by the genetic variability of that population. The more variability the faster the possible evolution; the less variability, the slower the evolution. He called this the Fundamental Theorem of Natural Selection. A corollary is a variable population has an increased chance of continuing to evolve and not going extinct when environmental conditions change.

One consequence of this is that populations long subjected to natural selection (in other words, all populations) would be expected to have little remaining variability for genes affecting fitness since such variability would have been diminished by selection.
If variability did not occur, all genotypes and phenotypes would be alike and there would be no differential survival or differential reproduction. In short, there would be no natural selection. This variability comes from mutation.