eduKateSG · Genetics · The Gold Standard of Education
Follow genes through space before calling a population locally adapted
Choose the explanation you need, or follow all twelve chapters in order.
Full chapter index · Independent practice · Genetics subject guide
Migration–selection balance describes the tension between gene flow and selection when environments differ across space. Migrants can carry alleles across a boundary; local selection can favour different alleles on each side. The result may be a stable frequency difference, a cline, a moving transition or no lasting contrast at all. None of those outcomes can be read from one allele-frequency map alone.
This article starts with a transparent two-deme example, then moves to clines, reciprocal transplants, genomic evidence and management limits. The examples are invented. The aim is to separate movement, reproduction and survival so that local adaptation is demonstrated rather than inferred from geography.
Continue from The Gold Standard of Genetics within The Gold Standard of Education: learning that transfers into life and the How X Works learning library.
Full chapter index
See the mechanism · Chapters 1–2
Build the worked model · Chapters 3–5
Test what changes · Chapters 6–8
Check evidence and limits · Chapters 9–10
Practise and continue · Chapters 11–12
CHAPTER 1 OF 12 · See the mechanism
1. Define migration, gene flow and local adaptation separately
Migration is movement of organisms or propagules. Gene flow occurs only when migrants reproduce and transmit alleles into the receiving population. A tagged fish can cross a river reach without contributing offspring; its movement is real, but its realised genetic contribution may be zero.
Local adaptation is a fitness pattern, not simply a genetic difference. In its strongest comparative form, a local population performs better in its home environment than foreign populations do there, and environments favour different populations. Neutral drift, barriers or shared history can create spatial allele-frequency differences without that pattern.
Declare the unit of analysis: deme, habitat patch, breeding season, life stage and genomic region. Also distinguish immigrant ancestry from contemporary migrants. These choices determine whether the evidence concerns movement, effective gene flow, selection or historical structure.
Evidence check: for Define migration, gene flow and local adaptation separately, preserve the observation before the interpretation. Record migration, gene flow and local adaptation as separate claims. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 2 OF 12 · See the mechanism
2. Draw the life cycle before writing an equilibrium
Order matters. In a migration-before-selection model, immigrants arrive and then experience local survival or reproduction. In a selection-before-migration model, local selection acts first and migrants are added later. The same symbols can produce different recursions when the biological order changes.
Write a cycle such as adults → gametes → dispersal → viability selection → breeding adults. Mark which stage is sampled. Juveniles collected after dispersal but before mortality may show abundant immigrant alleles even when those alleles contribute little to the next breeding cohort.
Migration–selection balance is an equilibrium statement: expected frequency change is zero under a specified, repeating model. A single year near that frequency does not prove equilibrium. Time series, demographic rates and fitness evidence are needed to show that opposing flows are actually balancing.
Evidence check: for Draw the life cycle before writing an equilibrium, preserve the observation before the interpretation. Record life-cycle order and sampled stage. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
Consider allele A in two equally sized demes. In habitat 1 it has relative fitness 1+s; the alternative allele has fitness 1. Let p be A’s frequency before selection. Under a simple haploid viability model, selection changes it to p_s = p(1+s)/(1+sp).
Suppose a fraction m of the next gene pool comes from deme 2, where A has frequency p₂. After selection in deme 1 and migration from deme 2, one simplified recurrence is p’ = (1−m)p_s + mp₂. This is a teaching model, not a universal biological law: diploidy, dominance, asymmetric movement and density regulation require different terms.
Keep each intermediate value. It shows whether selection raised A, whether incoming genes then diluted it, and how large the net change was. Compressing the whole life cycle into a final frequency hides the mechanism that must be tested.
Evidence check: for Build a one-locus, two-deme model, preserve the observation before the interpretation. Record one-locus two-deme recurrence. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 4 OF 12 · Build the worked model
4. Work the arithmetic and change one condition
Let p=0.70, s=0.20, m=0.10 and p₂=0.20. Selection gives p_s = 0.70×1.20/(1+0.20×0.70) = 0.84/1.14 ≈ 0.737. Migration then gives p’ = 0.90×0.737 + 0.10×0.20 ≈ 0.683. Selection raised A locally, but gene flow from the other habitat more than offset that increase in this generation.
Now reduce migration to m=0.02 while holding the other values fixed. Then p’ ≈ 0.98×0.737 + 0.02×0.20 ≈ 0.726. The allele still ends below its post-selection frequency because immigrants carry less A, yet it remains above the starting 0.70. The direction of one-generation change has reversed because the migration term changed.
Do not call 0.683 or 0.726 the equilibrium without iterating both demes under their own selection regimes and checking stability. The worked calculation is a mechanism audit: it reveals the sign and scale of competing forces under declared assumptions.
| Case | Starting p | After selection | After migration | Net change |
|---|---|---|---|---|
| m=0.10, source p=0.20 | 0.700 | 0.737 | 0.683 | −0.017 |
| m=0.02, source p=0.20 | 0.700 | 0.737 | 0.726 | +0.026 |
| m=0.10, source p=0.80 | 0.700 | 0.737 | 0.743 | +0.043 |
Evidence check: for Work the arithmetic and change one condition, preserve the observation before the interpretation. Record every intermediate value in the worked example. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 5 OF 12 · Build the worked model
5. Read a cline as a process-shaped transition
A cline is a spatial change in a trait or allele frequency. Its centre, width, slope and tails may contain information about dispersal, selection, environmental gradients and historical contact. A narrow transition can reflect strong selection, short dispersal or a barrier; it does not identify which cause by itself.
In simple tension or environmental cline models, greater dispersal tends to broaden the transition, while stronger opposing selection tends to narrow it. The exact relationship depends on the model, dimensionality, density and how fitness changes across space. Fit the biological model that matches the landscape rather than treating a smooth logistic curve as a complete explanation.
Sample beyond the apparent centre and repeat through time. Sparse sites can make a stepped boundary look smooth, while one storm, cohort or hybrid-zone movement can shift the estimated centre. Preserve coordinates, dates, life stages and sampling effort beside every frequency.
Evidence check: for Read a cline as a process-shaped transition, preserve the observation before the interpretation. Record cline centre, width and sampling design. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 6 OF 12 · Test what changes
6. Distinguish neutral structure from selected clines
Neutral markers describe shared history, drift and gene flow more directly than loci under divergent selection, but they are not a perfect control. Background selection, linked selection, mutation-rate differences and uneven marker ascertainment can affect them.
Compare candidate loci with a genome-wide background and explicit demographic models. If many unlinked markers change at the same geographic barrier, restricted movement or historical separation is plausible. If a small set of loci tracks the environment more sharply than neutral structure, divergent selection becomes more credible.
Population structure can also create spurious environment–allele associations because ancestry and habitat covary. Mixed models, matched sampling, simulations and independent populations help, but none substitutes for phenotype and fitness evidence.
Evidence check: for Distinguish neutral structure from selected clines, preserve the observation before the interpretation. Record neutral structure and linked-selection alternatives. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 7 OF 12 · Test what changes
7. Demonstrate local adaptation with fitness comparisons
Reciprocal transplant experiments move populations into each other’s environments and measure components of fitness such as survival, mating success or viable offspring. Common-garden experiments hold the environment more constant to test whether population differences persist.
Measure the whole relevant life cycle when possible. A genotype may germinate well yet reproduce poorly, or survive drought but lose mating opportunities. Short experiments can miss delayed costs and genotype-by-year interactions.
Replication across sites and years matters because “home” is not one immutable condition. A convincing claim states the populations, environments, response measure and interval tested. It avoids turning a context-bounded performance pattern into a claim of universal superiority.
Evidence check: for Demonstrate local adaptation with fitness comparisons, preserve the observation before the interpretation. Record home–away fitness evidence. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 8 OF 12 · Test what changes
8. Connect genomic, environmental and demographic evidence
Build three evidence layers. Genomic data estimate ancestry, differentiation and candidate regions. Environmental data describe the selective contrast at biologically relevant scales. Demographic data measure movement, survival and reproduction.
Agreement across layers is stronger than an extreme statistic alone. A candidate allele may track salinity, alter a measured physiological trait and predict reproductive success after controlling for ancestry. Each link still needs uncertainty and alternative explanations.
Linkage complicates interpretation because a selected allele can reduce effective gene flow at nearby neutral loci. Recombination can break that association, while low-recombination regions preserve broad blocks. Report genomic position and recombination information before counting neighbouring markers as independent evidence.
| Evidence layer | Primary observation | Supports | Main alternative |
|---|---|---|---|
| Genomic | Allele and ancestry patterns | Differentiation and candidate regions | Neutral structure or linked selection |
| Environmental | Exposure measured at biological scale | Selective contrast | Unmeasured correlated gradient |
| Demographic | Movement and realised reproduction | Effective gene flow | Detection or parentage bias |
| Fitness experiment | Home–away performance | Local adaptation | Year or maternal effects |
Evidence check: for Connect genomic, environmental and demographic evidence, preserve the observation before the interpretation. Record genomic, environmental and demographic agreement. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 9 OF 12 · Check evidence and limits
9. Test asymmetry, drift and changing environments
Real migration is often directional. Prevailing winds, currents, elevation or source–sink demography can send many more migrants one way. Unequal deme sizes also make the same migrant count a different fraction of each gene pool.
Drift is especially important in small demes. Allele frequencies can wander around deterministic expectations, and local extinction followed by recolonisation can mimic a sharp selection-driven turnover. Replicate populations and stochastic simulations show how much pattern the demographic process can generate alone.
Selection itself changes when temperature, competitors or land use change. A cline may lag behind a moving environmental boundary. Analyse time explicitly instead of forcing a stationary equilibrium on a transient system.
Evidence check: for Test asymmetry, drift and changing environments, preserve the observation before the interpretation. Record asymmetry, drift and temporal change. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 10 OF 12 · Check evidence and limits
10. Know the limits and common misconceptions
Gene flow is not always a force that prevents adaptation. It can introduce useful variation, rescue small populations or spread beneficial alleles. Its effect depends on direction, amount, ancestry, linkage and the receiving environment.
A steep cline does not prove strong selection, and a shallow cline does not prove weak selection. Barriers, recent contact, dispersal tails, density gradients and sampling design can produce similar shapes. Likewise, high FST is not a direct fitness measurement.
Management interventions such as assisted gene flow require more than a migration–selection calculation. They need disease, outbreeding, demographic, legal and community considerations plus monitoring and stopping rules. This article explains inference, not a universal translocation prescription.
Evidence check: for Know the limits and common misconceptions, preserve the observation before the interpretation. Record limits of equilibrium and FST interpretations. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
CHAPTER 11 OF 12 · Practise and continue
11. Independent practice and changed-context transfer
Practice case: in habitat 1, allele A begins at 0.60, selection raises its post-selection frequency to 0.66, and ten per cent of the next gene pool comes from a source where A=0.10. Calculate p’ and state whether the one-generation change supports or opposes A. Then repeat with a source where A=0.80.
Changed context: keep the observed cline fixed but add a newly discovered ridge that halves movement at the centre. Explain how the same width could now imply a different selection strength. Name one tagging, pedigree or genomic observation that would distinguish restricted dispersal from strong divergent selection.
Finish with a bounded claim containing the life-cycle order, spatial scale, time period, fitness measure and strongest unresolved alternative. If one of those is missing, the inference is not yet ready to travel to another population.
Evidence check: for Independent practice and changed-context transfer, preserve the observation before the interpretation. Record the changed source frequency that reverses direction. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.
Does migration always erase local adaptation?
No. Gene flow can oppose, maintain or sometimes facilitate adaptation depending on its rate, direction, genetic content and the strength and form of selection.
Is a cline an equilibrium?
Not automatically. A cline can move, broaden or reflect recent contact. Equilibrium must be evaluated with a time series and an appropriate process model.
What is the strongest evidence for local adaptation?
Fitness comparisons across home and away environments, supported by demographic, environmental and genomic evidence, are stronger than allele-frequency differences alone.
Continue with the Wahlund effect and hidden population structure, heritability, selection response and environment, and effective population size, bottlenecks and coalescent times.
Authoritative background includes the peer-reviewed two-deme analysis of gene flow and local adaptation, research on linkage, gene flow and local adaptation, and the empirical review causal agents and traits in local adaptation. Sources and links were checked on 22 October 2026.
Evidence check: for FAQ, sources and connected reading, preserve the observation before the interpretation. Record the narrowest local-adaptation claim supported. Then state the inference, one credible alternative, and the additional observation that would distinguish them. This compact audit keeps a fluent explanation tied to evidence that another reader can inspect.
Changed-context transfer: repeat this chapter’s decision after changing one consequential condition while holding the others fixed. Calculate or predict what should change, name the threshold or mechanism, and compare with the original case. If the conclusion survives, state the range tested; if it reverses, explain why. Transfer is demonstrated by the changed case, not asserted from the first example.
Practical decision record: write one sentence for what is known, one for what remains uncertain, and one for the next safe action. Include the evidence owner, the latest useful decision time, and the condition that would stop or reverse the action. This turns a technical result into a reviewable choice without pretending uncertainty has disappeared.