Natural selection and genetic drift
Two processes that change allele frequencies
- Mutation creates variation. Natural selection and genetic drift determine what happens to it.
- They are both mechanisms of evolution, but they differ in one decisive respect:
- Natural selection is non-random. Which alleles increase depends on their effect on survival and reproduction.
- Genetic drift is random. Which alleles increase depends on chance alone, regardless of their effect.
Natural selection
-
Natural selection follows from four observations, and setting them out in order is the clearest way to answer a "how" question:
- Variation — individuals in a population differ, and much of the difference is heritable.
- Overproduction — more offspring are produced than can survive on the available resources.
- Differential survival and reproduction — individuals whose characteristics suit the environment survive and reproduce more.
- Inheritance — those characteristics are passed on, so the frequency of the alleles responsible increases in the next generation.
-
Fitness in biology means reproductive success — the number of offspring surviving to reproduce. It does not mean strength, speed or health, except insofar as those affect reproduction.
Three patterns of selection
| Type | What is favoured | Effect on the population |
|---|---|---|
| Directional | One extreme | The mean shifts toward that extreme |
| Stabilising | The intermediate | Variation narrows; mean unchanged |
| Disruptive | Both extremes, against the intermediate | Variation increases; may split into two groups |
- Directional selection occurs when the environment changes or a population enters a new one — insecticide resistance is the classic case.
- Stabilising selection is the most common in a stable environment. Human birth weight is the standard example: very small and very large babies both historically had lower survival, so the intermediate was favoured.
- Disruptive selection matters most for this standard, because by favouring both extremes and selecting against intermediates it can begin to split a population — the first step toward sympatric speciation.
Genetic drift
-
Genetic drift is a change in allele frequency due to chance — random sampling of which individuals happen to survive and reproduce.
-
It occurs in all populations, but its effect depends critically on population size:
- In a large population, chance deviations in one direction are offset by deviations in the other, so frequencies stay close to expectation.
- In a small population, a few chance events can change frequencies dramatically, because each individual represents a large proportion of the whole.
-
Drift is not selection. An allele can be lost through drift even if it is beneficial, and can become common even if it is slightly harmful, purely by chance.
-
Two situations make drift especially powerful, and both are named in exams:
-
The founder effect.
- A small group leaves the main population and founds a new one.
- The founders carry only a sample of the original population's alleles, and by chance that sample is unrepresentative.
- Some alleles are absent entirely; others are at much higher frequency than in the source population.
- The new population therefore starts out genetically different from the parent population and with less variation — before selection has done anything at all.
-
The population bottleneck.
- A population is drastically reduced by a catastrophe — disease, habitat loss, predation, volcanic eruption.
- The survivors are a small, random sample, so much variation is lost.
- Even if numbers recover, the variation does not, because mutation restores it only very slowly.
Why this matters for speciation
-
Both processes cause populations to diverge, but by different routes:
- Selection drives divergence where two populations experience different environments, because different alleles are favoured in each.
- Drift drives divergence even in identical environments, because chance takes the two populations in different directions.
-
Divergence by either route can accumulate until the populations are reproductively isolated — which is speciation.
-
New Zealand is full of founder effects, because most of our fauna arrived as small numbers of colonists crossing ocean, and many species were then further divided among offshore islands.
Selective advantage and the two processes
- Under selection, the advantage is direct: individuals with the favoured allele leave more offspring, so the allele's frequency rises.
- Under drift, there is no advantage — that is the whole point. The frequency changes without any relationship to the allele's effect on survival. Answers that try to give a drift outcome an adaptive explanation have misidentified the process.
Worked Example
Worked Example
A native beetle occurs on the mainland and on a small offshore island colonised by a few individuals after a storm about 200 years ago.
- Mainland population: about 2 million beetles. Wing-colour alleles: dark 0.70, pale 0.30.
- Island population: about 800 beetles. Wing-colour alleles: dark 0.05, pale 0.95.
- The island and mainland habitats are similar in vegetation, predators and climate.
- The island population shows much lower genetic variation across all genes examined, not only wing colour.
- Pale beetles on both islands and mainland are taken by predators slightly more often than dark ones.
Explain the most likely cause of the difference between the two populations.
Answer:
Ruling out natural selection.
The obvious first hypothesis is that pale colouring is favoured on the island. The data rule this out on two grounds:
- The habitats are similar in vegetation, predators and climate, so there is no evident difference in selection pressure that would favour pale beetles there.
- Pale beetles are taken by predators slightly more often in both places. So selection acts against the pale allele on the island too — yet the pale allele is at 0.95. Selection cannot explain an allele reaching high frequency while being selected against.
Identifying the actual cause: the founder effect.
The island was colonised by a few individuals after a storm. Those founders carried only a small sample of the mainland population's alleles, and a small sample is very likely to be unrepresentative by chance.
If, by chance, most of the founding beetles happened to carry the pale allele, the new population would begin with a pale frequency far above the mainland's 0.30 — regardless of the allele's effect on survival. This is genetic drift acting through the founder effect.
Why the effect persisted. With only about 800 beetles, the island population remains small, so drift continues to be strong there. Selection against the pale allele is described as slight, and in a small population a weak selection pressure is easily overwhelmed by random sampling from generation to generation. On the mainland, with 2 million beetles, drift is negligible and even weak selection has kept the pale allele down at 0.30.
The decisive comparison is therefore not between the habitats but between the population sizes.
The confirming evidence.
The island population shows lower genetic variation across all genes examined, not just wing colour. This is important because:
- Selection acts on specific genes affecting a trait under pressure. It would reduce variation at the wing-colour locus but would not systematically reduce variation at unrelated genes.
- Drift and the founder effect affect the whole genome, because the founders carried a limited sample of every gene.
A genome-wide reduction in variation is therefore the signature of a founder effect, and it is what distinguishes the two hypotheses.
Why this matters for speciation.
The island population is now genetically different from the mainland one, and this happened without any difference in environment and without adaptation. Because the populations are separated by ocean, there is little or no gene flow between them, so the difference will persist and further differences will accumulate through continued drift and any local selection.
If this continues long enough, the populations may diverge until they are reproductively isolated — allopatric speciation initiated by chance rather than by adaptation. This shows that speciation does not require the two environments to differ.