5.17, 5.19 - Mutation, selection and speciation

Gene mutation supplies new alleles

A new allele begins with a change in DNA, rather than an organism deciding which characteristic it needs.

A substitution, insertion or deletion can change a gene's sequence. Some changes affect the amino acid sequence or amount of a protein, altering a phenotype. Others have no detectable effect on the phenotype. A mutation can therefore be advantageous, disadvantageous or effectively neutral in a particular environment; being new does not make an allele beneficial.

For a mutation to contribute to evolution across generations, it must enter a lineage that produces descendants. In sexually reproducing animals, a mutation in an egg, sperm or their precursor cells can be transmitted to offspring. A mutation confined to a skin cell is normally inherited only by that cell's descendants within the body, not by the animal's offspring. In an asexually dividing microorganism, a mutation can pass directly to daughter cells.

Mutations arise without regard to whether their effects will meet an organism's needs. DNA copying errors can occur before an environmental change, so a useful allele may already be present when selection begins. A new mutation can also arise later. Mutagens and cellular conditions can affect mutation rates, but that does not mean a population produces the particular useful change on demand. The classic Luria–Delbrück research distinguished the origin of resistant variants from their subsequent selection in bacteria.

Meiosis and fertilisation can generate new combinations of existing alleles. This changes offspring genotypes, but does not by itself create a new allele sequence. Similarly, producing more of an existing enzyme after a temperature change can alter an individual's phenotype without changing its inherited allele. Separate the origin of a DNA variant, the combination of variants in a genotype, and the expression of a phenotype.

Selection changes reproductive contribution

Natural selection changes how strongly different inherited variants contribute to the next generation.

Suppose an original, fictional population of haploid algae contains alleles $R$ and $r$ of one gene. Haploid means that each cell has one copy of this gene, so each cell carries either $R$ or $r$. The cells reproduce asexually. A previous gene mutation produced $R$, and both variants are present before the water becomes warmer.

In this model, $R$ produces a protein variant that maintains an essential cell process better at the new temperature. Its carriers can consequently produce more surviving daughter cells. The causal chain is:

1. Heritable allelic variation exists before selection acts on it.
2. At the higher temperature, the associated phenotype gives $R$ cells greater reproductive success than $r$ cells.
3. Daughter cells inherit their parent's allele, assuming no further mutation in this example.
4. The $R$ lineages contribute a larger share of the new cells, increasing the frequency of $R$ in the descendant generation.

Selection acts through phenotypic differences among individuals, but the evolutionary outcome is measured in the population's alleles. Individual $r$ cells do not have to become $R$ cells. Surviving longer matters to selection only insofar as it changes reproductive contribution: an organism that survives but leaves no descendants passes on none of its alleles through that route. Differential fertility or mating success can also produce selection without a difference in adult survival.

Fitness concerns relative reproductive success in the stated environment, rather than strength or a universal ranking of alleles. The $R$ protein might be costly at a lower temperature, or the alleles might perform equally there. Selection can change direction when conditions change. A beneficial allele can also be lost by chance, particularly while rare; selection does not guarantee its fixation, meaning a frequency of 1.

The distinction matters when describing climate responses: moving to a cooler place or changing enzyme expression during a lifetime is different from an inherited allele becoming more frequent over generations.

Counting allele change from descendants

Complete allele and descendant counts let us test whether a population has changed, while keeping the denominator explicit.

The following are original illustrative data for the haploid algal model. The first count is a complete census of the founding cells. Each founder's lineage is tracked through one asexual reproductive cycle, and the second count includes all established daughter cells at the same defined census stage. In this stipulated example there is no immigration, emigration or further mutation; the two alleles are the only variants of this gene. These are teaching data, not observations of a real algal species.

The founder ratio $R:r$ is $40:160=1:4$. The descendant ratio is $160:240=2:3$. Ratios compare the two categories; they do not use the same denominator as a frequency. In $1:4$, $R$ represents one of five total parts, not one of four.

The founding $R$ frequency is $40/200=1/5=0.20=20\%$. The descendant $R$ frequency is $160/400=2/5=0.40=40\%$. The frequency doubled even though the absolute number of $r$ cells also increased. Selection need not make every disfavoured lineage decline in number; its relative contribution can fall in an expanding population.

Here, $R$ lineages produced 4.0 established daughters per founder compared with 1.5 for $r$, a reproductive-output ratio of $4.0:1.5=8:3$. Those outcomes show differential reproductive contribution in the stipulated system and explain the change. Comparing 160 daughters with 240 daughters alone would miss the unequal numbers of founders.

Percentage points and relative percentage change

The increase from 20% to 40% is 20 percentage points: subtract the percentages. Its relative percentage increase is

The relative calculation uses the initial frequency as its denominator. It does not say that 100% of descendants carry $R$: only 40% do. Never use the fourfold rise in $R$ cell number, from 40 to 160, as if it were the change in allele frequency; the total population also changed.

With complete diploid genotype data, count allele copies directly: an $RR$ individual contributes two $R$ copies, $Rr$ one and $rr$ none, out of twice the total number of individuals. No Hardy–Weinberg assumption is needed for direct counting. A dominant-phenotype percentage alone does not give the allele frequency because homozygotes and heterozygotes can share the phenotype.

Allopatric speciation after geographical separation

A barrier can let populations follow separate evolutionary histories by reducing the exchange of alleles between them.

A beetle that crosses a river but leaves no descendants has moved without contributing alleles to the next generation there. Gene flow tends to mix alleles and can oppose divergence between populations. Reproductive isolation occurs when barriers reduce or prevent successful interbreeding and hence effective gene flow.

Allopatric speciation is the formation of new species following geographical isolation. Consider a river changing course and dividing a population of flightless beetles. If the new channel prevents them meeting and breeding across it, gene flow falls. Whether the river is a barrier depends on the organism: the same channel might barely restrict a flying insect.

The barrier does not immediately create new alleles or new species. Both sides can initially retain much of the ancestral variation. Subsequently, mutations arise independently, and different local conditions may favour different inherited variants through natural selection. Chance changes in allele frequencies, genetic drift, can also contribute, especially in small isolated populations. Reduced gene flow makes these differences less likely to be continually mixed away.