11.5: Selection

Natural selection drives adaptive evolution by selecting for and increasing the occurrence of beneficial traits in a population.

Learning Objectives

• Explain how natural selection leads to adaptive evolution

Key Points

• Natural selection increases or decreases biological traits within a population, thereby selecting for individuals with greater evolutionary fitness.
• An individual with a high evolutionary fitness will provide more beneficial contributions to the gene pool of the next generation.
• Relative fitness, which compares an organism’s fitness to the others in the population, allows researchers to establish how a population may evolve by determining which individuals are contributing additional offspring to the next generation.
• Stabilizing selection, directional selection, diversifying selection, frequency -dependent selection, and sexual selection all contribute to the way natural selection can affect variation within a population.

Key Terms

• natural selection: a process in which individual organisms or phenotypes that possess favorable traits are more likely to survive and reproduce
• fecundity: number, rate, or capacity of offspring production
• Darwinian fitness: the average contribution to the gene pool of the next generation that is made by an average individual of the specified genotype or phenotype

An Introduction to Adaptive Evolution

Natural selection only acts on the population’s heritable traits: selecting for beneficial alleles and, thus, increasing their frequency in the population, while selecting against deleterious alleles and, thereby, decreasing their frequency. This process is known as adaptive evolution. Natural selection does not act on individual alleles, however, but on entire organisms. An individual may carry a very beneficial genotype with a resulting phenotype that, for example, increases the ability to reproduce ( fecundity ), but if that same individual also carries an allele that results in a fatal childhood disease, that fecundity phenotype will not be passed on to the next generation because the individual will not live to reach reproductive age. Natural selection acts at the level of the individual; it selects for individuals with greater contributions to the gene pool of the next generation, known as an organism’s evolutionary fitness (or Darwinian fitness).

Figure \(\PageIndex{1}\): Adaptive evolution in finches: Through natural selection, a population of finches evolved into three separate species by adapting to several difference selection pressures. Each of the three modern finches has a beak adapted to its life history and diet.

Fitness is often quantifiable and is measured by scientists in the field. However, it is not the absolute fitness of an individual that counts, but rather how it compares to the other organisms in the population. This concept, called relative fitness, allows researchers to determine which individuals are contributing additional offspring to the next generation and, thus, how the population might evolve.

There are several ways selection can affect population variation:

• stabilizing selection
• directional selection
• diversifying selection
• frequency-dependent selection
• sexual selection

As natural selection influences the allele frequencies in a population, individuals can either become more or less genetically similar and the phenotypes displayed can become more similar or more disparate. In the end, natural selection cannot produce perfect organisms from scratch, it can only generate populations that are better adapted to survive and successfully reproduce in their environments through the aforementioned selections.

Galápagos with David Attenborough: Two hundred years after Charles Darwin set foot on the shores of the Galápagos Islands, David Attenborough travels to this wild and mysterious archipelago. Amongst the flora and fauna of these enchanted volcanic islands, Darwin formulated his groundbreaking theories on evolution. Journey with Attenborough to explore how life on the islands has continued to evolve in biological isolation, and how the ever-changing volcanic landscape has given birth to species and sub-species that exist nowhere else in the world.

Stabilizing, directional, and diversifying selection either decrease, shift, or increase the genetic variance of a population.

Learning Objectives

• Contrast stabilizing selection, directional selection, and diversifying selection.

Key Points

• Stabilizing selection results in a decrease of a population ‘s genetic variance when natural selection favors an average phenotype and selects against extreme variations.
• In directional selection, a population’s genetic variance shifts toward a new phenotype when exposed to environmental changes.
• Diversifying or disruptive selection increases genetic variance when natural selection selects for two or more extreme phenotypes that each have specific advantages.
• In diversifying or disruptive selection, average or intermediate phenotypes are often less fit than either extreme phenotype and are unlikely to feature prominently in a population.

Key Terms

• directional selection: a mode of natural selection in which a single phenotype is favored, causing the allele frequency to continuously shift in one direction
• disruptive selection: (or diversifying selection) a mode of natural selection in which extreme values for a trait are favored over intermediate values
• stabilizing selection: a type of natural selection in which genetic diversity decreases as the population stabilizes on a particular trait value

Stabilizing Selection

If natural selection favors an average phenotype by selecting against extreme variation, the population will undergo stabilizing selection. For example, in a population of mice that live in the woods, natural selection will tend to favor individuals that best blend in with the forest floor and are less likely to be spotted by predators. Assuming the ground is a fairly consistent shade of brown, those mice whose fur is most-closely matched to that color will most probably survive and reproduce, passing on their genes for their brown coat. Mice that carry alleles that make them slightly lighter or slightly darker will stand out against the ground and will more probably die from predation. As a result of this stabilizing selection, the population’s genetic variance will decrease.

Stabilizing selection: Stabilizing selection occurs when the population stabilizes on a particular trait value and genetic diversity decreases.

Directional Selection

When the environment changes, populations will often undergo directional selection, which selects for phenotypes at one end of the spectrum of existing variation.

A classic example of this type of selection is the evolution of the peppered moth in eighteenth- and nineteenth-century England. Prior to the Industrial Revolution, the moths were predominately light in color, which allowed them to blend in with the light-colored trees and lichens in their environment. As soot began spewing from factories, the trees darkened and the light-colored moths became easier for predatory birds to spot.

Directional selection: Directional selection occurs when a single phenotype is favored, causing the allele frequency to continuously shift in one direction.

Over time, the frequency of the melanic form of the moth increased because their darker coloration provided camouflage against the sooty tree; they had a higher survival rate in habitats affected by air pollution. Similarly, the hypothetical mouse population may evolve to take on a different coloration if their forest floor habitat changed. The result of this type of selection is a shift in the population’s genetic variance toward the new, fit phenotype.

Figure \(\PageIndex{1}\): The Evolution of the Peppered Moth: Typica and carbonaria morphs resting on the same tree.The light-colored typica (below the bark’s scar) is nearly invisible on this pollution-free tree, camouflaging it from predators.

Diversifying (or Disruptive) Selection

Sometimes natural selection can select for two or more distinct phenotypes that each have their advantages. In these cases, the intermediate phenotypes are often less fit than their extreme counterparts. Known as diversifying or disruptive selection, this is seen in many populations of animals that have multiple male mating strategies, such as lobsters. Large, dominant alpha males obtain mates by brute force, while small males can sneak in for furtive copulations with the females in an alpha male’s territory. In this case, both the alpha males and the “sneaking” males will be selected for, but medium-sized males, which cannot overtake the alpha males and are too big to sneak copulations, are selected against.

Diversifying (or disruptive) selection: Diversifying selection occurs when extreme values for a trait are favored over the intermediate values.This type of selection often drives speciation.

Diversifying selection can also occur when environmental changes favor individuals on either end of the phenotypic spectrum. Imagine a population of mice living at the beach where there is light-colored sand interspersed with patches of tall grass. In this scenario, light-colored mice that blend in with the sand would be favored, as well as dark-colored mice that can hide in the grass. Medium-colored mice, on the other hand, would not blend in with either the grass or the sand and, thus, would more probably be eaten by predators. The result of this type of selection is increased genetic variance as the population becomes more diverse.

Comparing Types of Natural Selection

Figure \(\PageIndex{1}\): Types of natural selection: Different types of natural selection can impact the distribution of phenotypes within a population.In (a) stabilizing selection, an average phenotype is favored.In (b) directional selection, a change in the environment shifts the spectrum of phenotypes observed.In (c) diversifying selection, two or more extreme phenotypes are selected for, while the average phenotype is selected against.

Fitness and Selection Coefficient

We can address the effects of selection quantitatively by asking what is the relative rates of reproduction of different genotypes? That is to say, how much of an advantage or disadvantage does a particular genotype confer? For example, let's consider two competing genotypes, \( A \) and \( B \) -- and let's further assume that genotype \( A \) is more fit. By convention, we assign the most fit genotype a relative fitness of 1 .... then all other genotypes have a relative fitness of less than 1. For example, perhaps genotype \( B \) has a relative fitness of \( 0.96 \) -- that means, for every offspring produced by genotype \( A \), genotype \( B \) produces (on average) \( 0.96 \) offspring.

A complementary way to talk about relative fitness is to ask how much of a fitness disadvantage does genotype \( B \) have relative to genotype \( A \)? We call this number the selection coefficient, and it is simply the complement of the relative fitness. If the relative fitness is \( w \) and the selection coefficient is \( s \), then

\[ s = 1 - w \]

Balancing Mutation and Selection

Selection operating alone will always drive the alleles contributing to the fittest genotypes to fixation (and the alleles contributing to less fit genotypes to loss.) However, selection is never operating alone! For example, consider a situation where a recessive allele causes a fitness disadvantage: ie, the fitnesses of genotypes \( AA \), \( Aa \) and \( aa \) are \(1\), \(1\), and \(1-s\) respectively. As usual, assume that \( p \) is the allele frequency of \( A \) and \( p \) is the allele frequency of \( a \). Because of directional selection, \( a \) alleles are lost over time at rate proportional to the genotype frequency. Assuming Hardy-Weinberg proportions, the genotype frequency of \( aa \) is \( q^2 \), so the rate of allele loss is \( q^{2}s \). However, \( a \) alleles are also being gained at a fixed rate \( \mu \). Thus, at equilibrium,

\[ q^{2}s = \mu \]

and thus the equilibrium allele frequency of \( a \) is

\[ \hat{q} = \sqrt{ \frac{\mu}{s} } \]

At realistic rates of selection and mutation, there is an important consequence of this equation: The equilibrium frequency of the deleterious allele is very low. As such, most of the deleterious alleles are carried by heterozygote carriers. The consequences of this are substantial, as we'll see next.

Partial dominance

What if the harmful allele \( a \) is somewhat dominant? That is, what if it shows a harmful effect on heterozygous carriers? Then, the fitness of genotypes genotypes \( AA \), \( Aa \) and \( aa \) are \(1\), \(1 - hs\), and \(1-s\) respectively. Parameter \(h\) is called the degree of dominance -- when \( h = 1 \), the harmfull allele is completely dominant. Alternatively, when \( h = 0 \), the harmful allele is completely recessive.

How does this change our analysis of the situation at equilibrium? Remember, most of the deleterious alleles are present in heterozygotes, and thus most of the alleles eliminated by selection are eliminated by selection against heterozygous genotypes. For each \( Aa \) genotype that doesn't reproduce, half of the lost alleles are \( a \). Recall also that the Hardy-Weinberg theorem tells us that the genotype frequency of \( Aa \) is \(2pq\). Thus, the number of alleles lost each generation is:

\[ 2pq * hs * 0.5 \]

As above, the number of deleterious alleles gained each generation is \( \mu \). At equilibrium, they are the same:

\[ 2pq * hs * 0.5 = \mu \]

Rearrange, and cancel the \(0.5\) and \(2\), to get:

\[ \hat{p}\hat{q}hs = \mu \]

AND ... remember how the harmful allele is rare at equilibrium? As a result, \( \hat{p} \) is so close to 1 that we can replace it in the above and solve for \{ \hat{q} \}:

\[ \hat{q} = \frac{\mu}{hs} \]

 

Heterozygote superiority

There's one more situation to address analytically -- that is when the heterozygote is superior. (This is the case, for example, in sub-Saharan Africa, where the allele that causes sickle cell anemia also gives heterozygotes resistance to malaria.) In this situation, the heterozygote genotype has a fitness of 1, and the two homozygotes have fitness that is less than 1. Let's assume that  the fitness of genotypes genotypes \( AA \), \( Aa \) and \( aa \) are \(1 - s\), \(1\), and \(1-t\) respectively. Recall that the genotype frequency of \( AA \) is \( p^2 \) and the genotype frequency of \( aa \) is \( q^2 \), Then, each generation, the proportion of \( A \) alleles eliminated by selection is \( \frac{p^{2}s}{p} \) or \( ps \), and similarly the proportion of \( a \) alleles eliminated by selection is \( \frac{q^{2}t}{q} \) or \( qt \). At equilibrium, these must be balanced, so \( \hat{p}s = \hat{q}t \), or:

\[ \hat{p} = \frac{t}{s + t} \] 

and similarly

\[ \hat{q} = \frac{s}{s + t} \]