Genetic Drift and Gene Flow

Natural selection is the headline mechanism of adaptive evolution, but it is not the only force changing allele frequencies in a population. Two complementary mechanisms — genetic drift (random sampling effects in finite populations) and gene flow (movement of alleles between populations) — operate alongside selection and, in particular circumstances, can dominate it. Drift is stochastic: it changes allele frequencies in directions unrelated to fitness. Gene flow is connective: it homogenises the genetic composition of subpopulations linked by migration. The interplay of these three forces (selection, drift, gene flow) plus a fourth (mutation) accounts for the full repertoire of microevolutionary change in real populations.

Spec mapping: This lesson sits in AQA 7402 Section 3.7.2 (genetic diversity and adaptation — effects of selection, genetic drift, and gene flow on allele frequencies) and feeds directly into Section 3.7.3 (speciation). Refer to the official AQA specification document for exact wording. The Hardy-Weinberg conditions of "large population" and "no migration" are the assumptions whose violation gives rise to drift and gene flow respectively (cross-link to lesson 1).

Connects to: Hardy-Weinberg equilibrium assumptions (Section 3.7.2, lesson 1); natural selection (Section 3.7.2, lesson 2); speciation (Section 3.7.3, lesson 4 — drift is a key driver of allopatric speciation).

---

What Is Genetic Drift?

Key Definition: Genetic drift is the random change in allele frequencies in a population due to chance variation in survival and reproductive output. It is not driven by differential fitness — the alleles that increase or decrease in frequency do so simply by stochastic sampling, not because they confer any advantage or disadvantage.

The mechanism is most easily understood by analogy with sampling. Imagine a population's gene pool as a large bag containing 50% red beads and 50% blue beads. If you draw 1,000 beads, you will get very close to 500 red and 500 blue — the sampling error is small relative to the sample size. If you draw only 10 beads, you might easily get 7 red and 3 blue; the sampling error is huge. The same logic applies to populations: the smaller the population, the larger the proportional sampling error in allele frequencies from one generation to the next.

Key Features of Genetic Drift

• Genetic drift affects all finite populations, but its effects are most significant in small populations (typically populations of fewer than a few hundred breeding individuals).
• In large populations, random fluctuations in opposite directions tend to cancel; in small populations, they can swing allele frequencies dramatically over a few generations.
• Drift can lead to fixation of an allele (frequency 1.0 — all individuals homozygous) or its loss (frequency 0.0). Once lost, an allele can be regenerated only by mutation or gene flow.
• Drift reduces within-population genetic diversity over time (because it tends to eliminate alleles by chance).
• Drift increases between-population divergence over time (because different drifting populations lose different alleles).
• Drift is random with respect to fitness — it can fix deleterious alleles and eliminate beneficial ones, especially when the selection coefficient s is small relative to 1/N (where N is population size).

Effective population size (N_e)

A subtle point — important at A-Level for the strongest answers — is that the population size relevant to drift is the effective population size (N_e), not the census population size. N_e measures the number of breeding individuals contributing to the next generation, weighted by reproductive variance. Skewed sex ratios, harem mating systems, or a small dominant minority of breeders all reduce N_e below the census count. In severely skewed mating systems (e.g. one alpha male siring most offspring), N_e can be a small fraction of the population. This means drift effects are stronger than naive census counts suggest.

---

The Bottleneck Effect

Key Definition: The bottleneck effect occurs when a population is drastically reduced in size by a catastrophic event (volcanic eruption, disease epidemic, hunting pressure, habitat destruction). The surviving population is a small, random sample of the original gene pool and may have substantially different allele frequencies.

How the bottleneck effect works

1. The pre-bottleneck population is large and contains many alleles at varied frequencies.
2. A catastrophe kills most of the population — typically non-selectively. Random survival means the survivors are a small, biased sample of the original gene pool.
3. The post-bottleneck population recovers in numbers but carries the allele frequencies of the small subset of survivors, not the original population.
4. Some alleles have been lost entirely; others have shifted dramatically in frequency.
5. Even though the population may return to its original size, its genetic diversity remains permanently reduced because the lost alleles cannot be recovered (except by mutation or gene flow from outside).

Documented examples

Northern elephant seals (Mirounga angustirostris). Nineteenth-century hunting reduced the population to a single breeding colony on Guadalupe Island, possibly as few as 20 individuals. Modern population numbers over 100,000, but molecular studies show extraordinarily low genetic diversity at allozyme loci and microsatellite markers compared to the closely-related southern elephant seal, which did not bottleneck. The low diversity persists today as a genetic legacy of that 19th-century crash.

Cheetahs (Acinonyx jubatus). Cheetahs experienced a severe bottleneck approximately 10,000–12,000 years ago, plausibly associated with the end-Pleistocene megafaunal extinctions. Modern cheetahs have such low genetic diversity that individuals from different geographic regions can accept reciprocal skin grafts without immune rejection — they are effectively immunological clones at MHC loci. Low genetic diversity is associated with poor sperm quality, low birth rates and elevated susceptibility to feline infectious peritonitis.

Vorlon-style worst case — humans. Modern humans show low effective population size by mammalian standards (N_e ≈ 10,000 for much of the Pleistocene), consistent with one or more historic bottlenecks during the out-of-Africa dispersal. By contrast, our closest extant relatives (chimpanzees) carry far higher genetic diversity, reflecting larger and more continuous historical population sizes.

---

The Founder Effect

Key Definition: The founder effect occurs when a small group of individuals colonises a new area and establishes a new population. The founders carry only a sub-sample of the alleles present in the original source population, so the new population's allele frequencies can differ markedly from the source's.

The founder effect is best thought of as a "displaced bottleneck": instead of the original population being reduced, a small fragment of it is reproductively isolated in a new location. The genetic consequences are similar.

How the founder effect works

1. A small group of individuals migrates from a large source population to a new habitat (e.g. an oceanic island, a remote valley, a religiously segregated community).
2. The founders carry only a subset of the source population's alleles. Rare alleles in the source may be over- or under-represented or absent altogether in the founders.
3. As the founding population grows, its allele frequencies reflect the founders, not the source.
4. Over generations, drift, selection and mutation continue to act, and the new population may diverge substantially from the source.

Documented examples

Amish community and Ellis-van Creveld syndrome. The Old Order Amish of Lancaster County, Pennsylvania, descend from a small group of approximately 200 German-speaking founders in the mid-18th century. By chance, at least one founder carried the recessive allele for Ellis-van Creveld syndrome (a form of dwarfism with polydactyly). Combined with later limited gene flow with the wider population (religious endogamy) and a small effective population size, the allele rose to a frequency far above its frequency in the source German population. Modern incidence is approximately 1 in 200 Amish births versus roughly 1 in 60,000 in the general European-descended population.

Pingelap Atoll and achromatopsia. In 1775, a typhoon and subsequent famine devastated Pingelap Atoll in Micronesia, leaving approximately 20 survivors. One of these survivors carried the recessive allele for complete achromatopsia (total colour blindness with reduced acuity). Today, approximately 5–10% of the island's population is affected — versus roughly 0.003% worldwide. This is the textbook example of a combined bottleneck (typhoon) and founder effect (small post-disaster population that became the founders of the modern population).

Galápagos giant tortoises. Each volcanic island in the Galápagos was colonised by a small founding group of tortoises drifting from a parental population (most likely on the South American mainland or another Galápagos island). Drift and divergent selection have since produced distinguishable subspecies on different islands — a textbook starting point for allopatric speciation (lesson 4).

---

Comparing Bottleneck and Founder Effects

Feature	Bottleneck effect	Founder effect
Trigger	Population drastically reduced by catastrophe	Small group colonises new area
Population affected	Existing population at original location	New population in new location
Mechanism	Random survival of small subset of pre-existing population	Random sample of alleles taken from source population
Effect on diversity	Reduced	Reduced (in the new population; source unaffected)
Allele-frequency change	Random with respect to fitness	Random with respect to fitness
Long-term consequence	Population recovers numerically but with reduced variation	New population becomes genetically distinct from source; potentially a precursor to speciation
Canonical example	Northern elephant seals, cheetahs	Amish Ellis-van Creveld, Pingelap achromatopsia

---

Gene Flow (Migration)

Key Definition: Gene flow is the movement of alleles between populations through the migration of individuals — or, in plants, the movement of gametes (pollen, seeds). Gene flow introduces new alleles into the recipient population and shifts allele frequencies toward those of the donor.

Effects of gene flow

• Gene flow increases genetic diversity within a recipient population by introducing alleles that may not have been present before.
• Gene flow reduces genetic differences between populations — it has a homogenising effect, making populations more similar to each other genetically.
• Gene flow can counteract the effects of natural selection and drift. If selection acts against an allele in one population but gene flow continually reintroduces it from a neighbouring population, the allele will persist at higher frequency than selection alone would predict.
• Conversely, if gene flow is restricted by geographical or behavioural barriers, populations can diverge genetically and may eventually become separate species. Restricting gene flow is therefore the enabling condition for allopatric speciation (lesson 4).

Quantifying gene flow

The amount of gene flow between two populations is typically measured by the migration rate (m) — the proportion of the recipient population in each generation that consists of immigrants. Even very low migration rates (m of order 0.01) are sufficient to prevent significant genetic divergence between populations through drift, because Nm (the number of migrants per generation) only needs to exceed roughly one to keep the populations genetically connected. This is why even isolated populations connected by occasional dispersers tend not to diverge into separate species; they remain part of a single gene pool.

Examples of gene flow

Wind-pollinated plants. Trees such as oaks (Quercus) and pines (Pinus) can disperse pollen over many kilometres on the wind, maintaining gene flow between populations that may be visually separate. This is why oak populations across Europe show relatively shallow genetic structure despite the absence of physical individual movement between them.

Animal dispersal. Many bird species disperse from natal sites in their first year, mixing alleles across continental ranges. Even occasional long-distance migrants (e.g. vagrant individuals blown off course) can introduce alleles into populations otherwise considered isolated.

Habitat fragmentation reduces gene flow. A new motorway through a forest can cut off populations on either side; over decades, drift in each fragment leads to allele-frequency divergence. Conservation biologists therefore design wildlife corridors specifically to maintain gene flow between fragmented populations and preserve effective population size.

---

The Interaction of Selection, Drift, and Gene Flow

In real populations, all three forces act simultaneously, plus mutation. The dominant force depends on context:

Force	Effect	Dominant when...
Natural selection	Directional change in allele frequencies toward higher fitness	Selection strong (s > 1/N_e), population large
Genetic drift	Random fluctuations in allele frequencies	Population small (N_e small relative to 1/s)
Gene flow	Homogenises allele frequencies between populations	Migration rate Nm > 1
Mutation	Introduces new alleles slowly	Always present; rarely the immediate driver of frequency change

flowchart TD
  A["Mutation: introduces new alleles"] --> E["Allele frequencies"]
  B["Selection: directional, fitness-dependent"] --> E
  C["Drift: random, dominates in small populations"] --> E
  D["Gene flow: homogenises across populations"] --> E
  E --> F["Population genetic structure"]
  F --> G["Speciation if isolation persists"]

Key principles

• In large populations, selection is usually the dominant force; drift has minimal effect because sampling errors average out.
• In small populations, drift can overpower selection — beneficial alleles may be lost by chance, deleterious alleles may be fixed by chance. This is Wright's threshold: selection effectively dominates only when 4N_e × s > 1.
• Gene flow connects populations and prevents them from diverging by drift or selection. Restricting gene flow is therefore the prerequisite for population divergence and speciation.
• The four-force model (mutation, selection, drift, gene flow) is the foundational framework of population genetics.

Exam Tip: When asked to explain why a rare allele has a high frequency in a particular population, the standard four candidates are: (i) founder effect, (ii) bottleneck effect, (iii) heterozygote advantage (from lesson 1), and (iv) drift in a small isolated population. Always link your answer to specific circumstances of the named population — the marker is looking for evidence that you can apply the correct mechanism to the scenario, not just list the four candidates.

---

A-Level-Depth Misconceptions

• Treating drift as a directional force. Drift is stochastic — it has no direction. Students sometimes write that drift "removes harmful alleles", but this confuses drift with selection. Drift can equally well fix harmful alleles or remove beneficial ones; what makes it drift, not selection, is that the direction is unrelated to fitness.
• Confusing the bottleneck effect with the founder effect. Both reduce N and reduce diversity, but the bottleneck applies to an existing population, while the founder effect creates a new population. The Pingelap case is both: the typhoon caused a bottleneck of the original population, and the survivors are also the founders of the modern population.
• Assuming drift only affects small isolated populations. Drift acts on all finite populations — only its magnitude depends on N_e. Even the human population is subject to drift at neutrally evolving loci.
• Treating gene flow as always equalising. Gene flow homogenises only if it is reciprocal or symmetrical. Asymmetric gene flow (mainland-to-island, with little island-to-mainland return) can shift the recipient population toward the source's allele frequencies without doing the reverse.
• Forgetting effective population size. A population of 100,000 with one dominant male and a 99,999-male skew has an N_e closer to 4 than to 100,000. Students who quote census counts when N_e is what matters lose marks at the A* boundary.

---

Required Practical Cross-Reference