Questions: Genetic Drift: Process and Population Effects

Question 1

A beneficial allele that increases fitness by 0.01% (s = 0.0001) arises as a single new mutation in a population of N = 1,000 individuals. What is the most likely fate of this allele?

• A. It will spread to fixation because it is beneficial and natural selection is deterministic
• B. It will be lost, because when it is rare its fate is dominated by drift (Ns ≈ 0.1 ≪ 1), making it behave nearly neutrally
• C. It will reach an intermediate stable frequency because selection and drift balance each other
• D. It will spread to fixation slowly, because selection always wins over drift in the long run

Answer: (B)

Whether selection or drift dominates depends on the product Ns. When Ns ≪ 1 (here Ns = 1000 × 0.0001 = 0.1), the allele behaves as effectively neutral: drift overwhelms the tiny selective advantage. As a single new mutation, it starts at frequency 1/2N = 1/2000. The probability of fixation for a neutral allele is simply 1/2N ≈ 0.05% — nearly certain loss. A selection coefficient of 0.01% cannot overcome the massive sampling variance in a population of only 1,000. Option D is the common misconception: selection does not 'always win in the long run' for mildly beneficial alleles in finite populations.

Question 2

In a population of 10,000 individuals, a neutral allele is currently at frequency 5% (p = 0.05). What is its probability of eventually fixing (reaching p = 1.0)?

• A. Essentially zero — neutral alleles are almost always lost because drift is too weak at this population size
• B. 5% — equal to its current frequency, because fixation probability of a neutral allele equals its current frequency
• C. 50% — because drift is symmetric, there is an equal chance of going up or down
• D. About 1/2N = 0.005% — because the fixation probability equals the initial frequency when the allele first appeared

Answer: (B)

For a neutral allele, the probability of fixation equals its current frequency p. This is a fundamental result of drift theory: if p = 0.05, there is a 5% chance this allele eventually goes to fixation, and a 95% chance it is lost. Importantly, this probability holds regardless of population size — the population size affects *how long* fixation takes (approximately 4N generations), not the probability itself. Option C (50%) is wrong; symmetry of drift means expected frequency change is zero, but fixation probability depends on starting frequency. Option D confuses the initial frequency at first appearance (1/2N) with the current frequency.

Question 3

Genetic drift operates in a predictable, directional manner, systematically pushing allele frequencies toward values that enhance population fitness.

• T. True
• F. False

Answer: False

False. Genetic drift is inherently random and undirected — it is sampling error, not a force with direction or tendency. In any given generation, drift is equally likely to increase or decrease an allele's frequency (the expected change is zero). Over time, allele frequencies perform a random walk that must eventually end in fixation or loss, but which outcome occurs for any particular allele is stochastic. Natural selection is the directional force that favors higher-fitness alleles; drift is the random noise around that signal. The misconception that drift has a direction often arises from conflating drift with the Founder Effect, where a small founding population may by chance carry unusual allele frequencies — but this is still a random sampling outcome, not a directed process.

Question 4

Genetic drift can cause a beneficial allele to be permanently lost from a population before selection has a chance to spread it.

• T. True
• F. False

Answer: True

True. When a beneficial allele first arises as a mutation, it exists at very low frequency. At low frequencies, the random component of its fate (drift) is large relative to the deterministic component (selection). Even an allele with a 10% fitness advantage has approximately a 20% chance of fixation when it first arises — meaning an ~80% chance of loss, primarily from drift while it is rare. This has important consequences: many beneficial mutations are lost before they spread, evolution is not a reliable optimizer, and the efficacy of selection depends critically on Ne. The neutral theory of molecular evolution partially rests on this observation: most alleles that fix are neutral, fixed by drift, because beneficial alleles are so often lost before they can spread.

Question 5

Why does drift become negligible relative to selection as population size increases, and what is the condition that determines whether an allele 'behaves as if neutral'?

Model answer: The variance in allele frequency change per generation due to drift is p(1-p)/2N — it decreases as N increases. Selection causes a deterministic frequency change of approximately sp(1-p) per generation. The ratio of selection to drift scales with Ns: when Ns ≫ 1, selection's deterministic push far exceeds drift's random fluctuations, and alleles behave as their fitness predicts. When Ns ≪ 1, drift's noise swamps selection's signal, and the allele behaves as effectively neutral regardless of its actual s. The boundary condition Ns ≈ 1 (equivalently, s ≈ 1/N) marks when selection and drift are approximately equal in strength.

This is the central quantitative insight of population genetics: it is not s alone, or N alone, but their product Ns that determines which force dominates. This has major implications: the same allele with s = 0.001 behaves as neutral in a population of 100 (Ns = 0.1) but is strongly selected in a population of 10,000 (Ns = 10). Conservation geneticists use this principle to assess the minimum viable population size needed for selection to function effectively.