A comparison
Founder effect vs bottleneck effect
Both are genetic drift in a population briefly reduced to a few individuals; the difference is what happens to everyone else.
On this page
The founder effect happens when a few individuals leave a large population and start a new one. The bottleneck effect happens when a population itself crashes to a few individuals and then recovers. Genetically they are the same event: the next generation descends from a small random sample of gene copies, so rare variants are usually lost, a few become common by accident, and diversity falls. The difference is the story around the sample. After a founding, the source population is still there, unchanged; after a bottleneck, there is no one else[1]. Both are forms of genetic drift.
The two side by side
| Founder effect | Bottleneck effect | |
|---|---|---|
| What happens | A few individuals start a new population elsewhere | An existing population crashes to a few, then recovers |
| The source population | Still exists, with its variation intact | Is the population: there is no source left |
| Genetic cause | Random sampling of a few gene copies: Binomial(2n, p) copies of an allele at frequency p | |
| Typical result | Rare alleles lost; some made common by chance; heterozygosity reduced | |
| Can the variation return? | Yes, by migration from the source | Only by mutation, or migration from elsewhere if any exists |
| Classic example | Ellis–van Creveld syndrome in one Amish community | Northern elephant seals; Pingelap |
The founder effect: a boatload is a small sample
If n founders leave a population where an allele has frequency p, the number of copies they carry is Binomial(2n, p). With ten founders and an allele at 1 %, the chance they carry none is 0.9920 ≈ 82 %; but in most of the remaining boatloads the allele starts at 5 %, five times its source frequency. Ellis–van Creveld syndrome, common in a single Amish community and rare elsewhere, is the textbook case[2][3]. documented
The bottleneck effect: diversity lost is not regained
A bottleneck applies the same sampling to the whole population, often for several generations. Each generation at size N removes a fraction 1/2N of the heterozygosity, so a few generations at a handful of individuals cost more than centuries at a thousand, and recovery in numbers does not restore what was lost[1]. Northern elephant seals, reduced to perhaps 20 animals by 1892, kept about one eighth of their genetic variation through a more than ten-thousandfold recovery[4]. On Pingelap a typhoon is thought to have left a few survivors, at least one carrying the allele for complete achromatopsia, which is now common there[5][6]. reconstructed
After either event: drift keeps going
A new population of a few dozen keeps drifting strongly for generations, whether it was founded or crashed. That can push a lucky allele further up or remove it altogether. The figure below follows thirty such populations.
The longer treatment, with the equations and both case studies in full, is Chapter 4. For more cases of drift, see examples of genetic drift.