Gene flow

Concept

Gene flow

T-048Home BU-204Threads information · evolution

Statement

Migration mixes alleles between populations.

Why it matters

hardy-weinberg establishes the idealised baseline in which allele frequencies never change; gene-flow is one of the concrete forces that, once present, drives real populations away from that baseline. Alongside genetic-drift and mutation-selection-balance, it is one of the standard mechanisms population geneticists check for when an observed population's allele frequencies deviate from Hardy-Weinberg predictions, and it plays a distinctive role among them: unlike drift (random) or mutation (rare and slow), gene flow actively moves existing alleles between populations and, given enough of it, actively works to homogenize them.

Understanding gene flow's magnitude is also directly relevant to conservation biology and to interpreting population structure: whether two populations are effectively a single interbreeding unit or two genetically diverging ones depends critically on how much gene flow connects them.

Hypotheses

Migrant individuals interbreed successfully with the recipient population.Gene flow specifically requires successful reproduction and allele transfer, not merely physical movement; an individual that relocates but fails to interbreed (behavioural, ecological, or reproductive barriers) contributes no gene flow at all, regardless of how many individuals physically move. The migrant pool's allele frequency differs from the recipient population's.If migrants are drawn from a population with an identical allele frequency to the recipient's, migration produces no frequency change even though individuals are moving and interbreeding freely — it is specifically the frequency difference between source and recipient that drives any effect.

Proof

p' = (1-m)p + m\,p_m

One-generation migrant-pool model: if a fraction \(m\) of the next generation's individuals derive from migrants (allele frequency \(p_m\)) and the remainder \((1-m)\) from the existing recipient population (frequency \(p\)), the new population-wide frequency is this weighted average. A

\Delta p = p'-p = m(p_m-p)

Rearranging Step 1 directly: the frequency shift in a single generation is proportional both to the migration rate \(m\) and to how different the migrant pool's frequency is from the recipient population's, matching the Hypotheses' requirement that both a nonzero \(m\) and a nonzero frequency difference are needed for any effect. A

\text{Repeated migration at constant rate } m \text{ each generation drives } p \text{ toward } p_m, \text{ with } (p_m-p) \text{ shrinking by a factor } (1-m) \text{ per generation.}

Applying Step 2 repeatedly shows the gap between recipient and source frequencies decays geometrically over successive generations, so sustained gene flow, even at a modest constant rate, eventually equalises frequencies between two connected populations if given sufficient time. B

\text{Because gene flow moves actual alleles between populations, even a small } m \text{ is generally sufficient to prevent populations from diverging under drift or local selection alone.}

Unlike genetic-drift (which reshuffles frequencies randomly within a single population without introducing new alleles from outside) and mutation (which creates novel variants at a typically far slower rate), gene flow directly counters divergence by continually reintroducing the other population's alleles, acting as a homogenising force across a species range wherever connectivity exists. A

\text{Genetic differentiation between populations, measured by statistics such as } F_{ST}, \text{ falls as sustained migration rate } m \text{ rises.}

Populations exchanging migrants at a high rate show low \(F_{ST}\) (little differentiation), while populations exchanging few or no migrants can diverge substantially under drift or local selection alone; \(F_{ST}\) is therefore a commonly used indirect measure of historical gene flow between populations. B

Result

\Delta p = m(p_m - p)

Reading. Gene flow shifts a recipient population's allele frequency toward the migrant source's frequency, at a rate set by the migration rate itself and the size of the frequency gap between the two populations.

Scope. Applies once migrants successfully interbreed (Hypotheses); has no effect if either migration itself is absent (\(m=0\)) or migrants happen to share the recipient's exact allele frequency (\(p_m=p\)), even with substantial physical movement of individuals.

Corollaries & converses

Gene flow is one of the standard forces (alongside genetic-drift, mutation-selection-balance, and selection itself) that individually violate a hardy-weinberg assumption; observing that a population deviates from Hardy-Weinberg predictions motivates checking each of these in turn as a possible explanation.
inbreeding-heterozygosity describes, in a sense, the opposite extreme — a restricted mating pool with effectively no incoming gene flow, concentrating existing variation rather than mixing it with an external source.
Converse: observing very low genetic differentiation (\(F_{ST}\)) between two geographically separated populations is itself indirect evidence of substantial historical gene flow between them, without needing to observe migration directly.

Fails without

Migrants fail to interbreed successfully (Hypotheses): physical relocation of individuals with no actual reproduction in the new population contributes nothing to allele frequency, regardless of how many individuals move; \(m\) in the formula is then effectively zero for genetic purposes even while dispersal itself is clearly occurring.
Migrant frequency equals recipient frequency (Hypotheses): if \(p_m=p\), then \(\Delta p=m(p_m-p)=0\) exactly by Step 2, regardless of how large \(m\) is — moving individuals between populations with identical allele frequencies produces no detectable genetic change even with very substantial migration.

Common errors

Equating physical dispersal or migration of individuals with gene flow itself; successful interbreeding is required (Hypotheses), and reproductive or behavioural isolation can decouple the two entirely.
Assuming gene flow is always detrimental to a recipient population's adaptation; it can also introduce beneficial variation from elsewhere (sometimes called a genetic rescue effect), not only diluting locally adapted alleles.
Treating migration rate \(m\) as a fixed, species-wide constant rather than something that varies by specific population pair, by generation, and often by season or environmental condition.
Confusing gene flow's homogenising, between-population effect with genetic drift's randomising, within-population effect; the two forces act in fundamentally different directions on genetic differentiation and are easily conflated.

Discussion

Sewall Wright's island model of population structure, developed in the early-to-mid 20th century alongside his broader work in population genetics, is the classic formal treatment of migration between a set of connected subpopulations exchanging genes at a specified rate; the simple one-generation model derived here (Step 1) is a direct simplification of that broader framework, isolating gene flow's effect from the other forces Wright's fuller models also incorporate.

In practice, real gene flow is rarely a single constant rate between two discrete populations; stepping-stone models (gene flow occurring preferentially between geographically adjacent populations, weaker between distant ones) and isolation-by-distance patterns (genetic differentiation increasing smoothly with geographic distance) are more realistic elaborations of the same underlying logic developed here.

Common misconception: that any amount of gene flow, however small, is sufficient to prevent meaningful genetic divergence between populations. While Step 3 shows that sustained gene flow does erode differentiation over time, sufficiently strong local selection or sufficiently strong drift (in a very small recipient population) can still maintain substantial divergence even against a low but nonzero migration rate; the outcome depends on the relative strength of gene flow against whatever opposing forces are simultaneously acting.

Worked examples

p=0.20\ (\text{recipient}), \quad p_m=0.60\ (\text{migrant source}), \quad m=0.05

Applying Step 2 of the Proof: \(\Delta p = m(p_m-p) = 0.05\times(0.60-0.20)=0.05\times0.40=0.020\). A single generation of migration at a 5% rate shifts the recipient population's allele frequency from 0.20 to 0.22. A

\text{Repeating for a second generation, using the new frequency } p'=0.22\text{:}

\(\Delta p_2 = 0.05\times(0.60-0.22)=0.05\times0.38=0.019\), giving \(p''=0.239\); each successive generation's shift is slightly smaller than the last, since the gap \((p_m-p)\) itself is shrinking (Step 3's geometric decay), even though \(m\) is held constant across generations. A

p:\ 0.20\to0.22\to0.239\to\cdots\to p_m=0.60\ (\text{asymptotically})

Reading. Even a modest, constant migration rate steadily and predictably shifts a recipient population's allele frequency toward the migrant source's frequency, with the rate of approach slowing as the gap between the two narrows.

Scope. The identical recursive calculation, applied generation by generation, predicts the full trajectory toward eventual equalisation for any starting \(p\), \(p_m\), and constant \(m\).

Problems

A population has allele frequency \(p=0.30\); each generation, 10% of its individuals are migrants from a source population with frequency \(p_m=0.80\). Compute \(\Delta p\) for the first generation of migration.
Solution
\(\Delta p = m(p_m-p) = 0.10\times(0.80-0.30) = 0.10\times0.50 = 0.050\). The recipient population's frequency rises from 0.30 to 0.35 after one generation.
Two populations show nearly identical allele frequencies at a given locus, and direct observation confirms very little physical movement of individuals between them. Propose an alternative explanation for the similar frequencies, other than high current gene flow.
Solution
Similar frequencies do not require high current gene flow; the two populations could instead share a recent common ancestral population (having only recently diverged, with insufficient time for drift or differential selection to produce noticeable divergence yet), or could be independently maintained at similar frequencies by parallel selective pressures acting on each population separately, without any migration connecting them at all.
Explain, using the Hypotheses, why building a physical barrier that prevents individuals from two adjacent populations from moving between them, without otherwise affecting the populations, would eliminate gene flow between them even if a small number of individuals could still occasionally cross.
Solution
Gene flow requires successful interbreeding, not merely the possibility of movement (Hypotheses); if the barrier is complete enough that essentially no individuals can cross and interbreed, \(m\) becomes effectively zero regardless of the underlying allele-frequency difference \(p_m-p\), and by Step 2 of the Proof, \(\Delta p = m(p_m-p) \to 0\) as \(m\to0\), so the two populations would be free to diverge under drift or local selection with no gene-flow force opposing that divergence.