Genetic Drift Explained: Chance, Bottlenecks and Founder Effects

Evolution is often pictured as a steady process in which the best-adapted variants win. Yet allele frequencies also change for a reason that has nothing to do with fitness: chance. Each generation is formed by a finite sample of gametes drawn from the parents, and by luck some alleles are passed on more often than others. This process is called genetic drift.

This article explains how drift works, why its strength depends on population size, and how bottlenecks and founder events can reshape a population's genetic makeup.

Sampling is the core idea

Imagine a large jar containing red and blue marbles in equal numbers. Pull out ten marbles at random, and the result may not be five red and five blue. Pull out ten thousand, and the ratio will almost always be very close to half and half. Small samples fluctuate more than large ones.

Genetic drift works the same way. Each generation, the alleles that become the next generation are a random sample of the alleles in the parents. A sample of two offspring carries very different odds than a sample of two million. The smaller the population, the larger the random swings in allele frequency.

Drift is not selection

Natural selection changes allele frequencies in a direction set by fitness: advantageous alleles tend to rise. Drift has no such direction. It can raise or lower any allele, including ones that help, harm, or do nothing at all.

This difference matters for interpretation. A change in allele frequency is not, by itself, evidence of adaptation. Population geneticists often need to ask whether observed patterns match what drift alone would produce, or whether they require selection to explain them.

Fixation and loss

Over many generations, drift tends to push each allele toward one of two endpoints. Either the allele is lost from the population, reaching a frequency of zero, or it becomes fixed, reaching a frequency of one, meaning every individual carries it. Fixation can occur even for an allele that is mildly harmful, and loss can occur for one that is beneficial, simply because of sampling luck.

This is why drift is described as a randomizing force. Given enough time, it removes genetic variation within a population, even if the population is not changing in size or environment.

The role of population size

The strength of drift depends on population size. In a population of N individuals, the expected rate at which genetic diversity is lost each generation is roughly 1 divided by 2N, for a diploid population of N individuals. So drift is weak when N is large and strong when N is small.

Population size	Typical drift strength	Consequence
Very large (millions)	Very weak	Selection dominates most variants
Moderate (thousands)	Moderate	Selection and drift both matter
Small (tens)	Strong	Chance can override mild selection
Very small (a few)	Very strong	Alleles lost or fixed quickly

The table gives the general pattern rather than exact thresholds. The key point is that the same allele can be ignored by drift in a huge population and dominate by chance in a tiny one.

Bottlenecks

A bottleneck is a sharp reduction in population size. Disease, habitat loss, hunting or natural disasters can shrink a population to a small number of survivors. The survivors are a random sample of the original gene pool, so many alleles present before the crash may be lost entirely, and common alleles may become much more frequent by chance.

When the population later recovers, its genetic diversity remains reduced. This matters for conservation, because a species with little genetic variation may have less capacity to adapt to new diseases or environmental change. Several species, including the northern elephant seal, passed through severe bottlenecks in the past and show reduced diversity today.

Founder effects

A founder effect occurs when a new population is started by a small group of colonists. The founders carry only a subset of the alleles in their source population. Whatever alleles they happen to carry are the ones that begin the new population, and those alleles may be far more or less common than they were in the original group.

Founder effects are common in island colonization, in isolated communities and in populations that split from a larger group. Some human isolated populations show elevated frequencies of rare inherited conditions for this reason, because a few founders carried those alleles and the population remained small for generations.

Measuring drift

Geneticists track drift through measures such as heterozygosity, the fraction of individuals carrying two different alleles at a locus. In an idealized population of constant size, heterozygosity declines by a factor of about 1 minus 1 divided by 2N each generation. Over time, this decline is a clear signature of drift, especially when observed across many loci in the same population.

The rate of this decline is one of the reasons effective population size, rather than the simple head count, is important in conservation. Effective size reflects how many individuals actually contribute genes to the next generation, which can be far lower than the total population when breeding is uneven.

Drift and selection together

Drift and selection are not mutually exclusive. In large populations, selection usually dominates, because even small fitness differences are enough to overcome sampling noise. In small populations, drift can overwhelm weak selection, allowing harmful alleles to spread or beneficial alleles to disappear.

This interplay helps explain why small, isolated populations often accumulate features that are not strongly adaptive. Their evolution is shaped more by chance than by the careful sorting of selection, and the results can look unusual or even maladaptive.

Common misunderstandings

Drift is sometimes described as a weaker form of selection, but this is misleading. It is a separate process with its own properties. It is also sometimes confused with random mutation. Mutation creates new alleles, while drift changes the frequencies of alleles that already exist. Both are random in some sense, but they play different roles.

Another misunderstanding is that drift only matters in tiny populations. In practice, drift shapes genetic variation in every population, but its effects are strongest where population sizes are small or where bottlenecks and founder events have occurred.

Why drift matters

Drift helps explain patterns in conservation genetics, human population history and the diversity of species on islands and in isolated habitats. It also provides a null model: when researchers see a gene variant that rises or falls, they often ask whether drift alone could account for it before concluding that selection was involved.

Understanding drift therefore sharpens the interpretation of evolutionary evidence. It reminds us that the genetic makeup of a population is partly a record of history, shaped by the accidents of who survived, who reproduced, and which gametes happened to meet.

Key takeaways

Genetic drift is the random change in allele frequencies that arises from sampling a finite number of gametes each generation. It has no direction tied to fitness, and it can lead alleles to be lost or fixed purely by chance.

Drift is strongest in small populations, and it is intensified by bottlenecks and founder effects. Together with selection, it shapes the genetic variation of every population, and it provides the baseline against which adaptive explanations must be tested.