Difficulty: Intermediate | Prerequisites: Allele and genotype frequencies, Hardy-Weinberg equilibrium, basic probability concepts (BIOL 180 Lectures 14-15; companion mutation notes from L16).
This is the second half of the lecture on non-selective mechanisms of evolution. Where the mutation notes covered a slow, directional force, genetic drift is random and can be fast, especially in small populations. Drift is the reason that small, isolated populations behave so differently from large ones, and it underpins concepts you will meet later in the course, including the founder effect and loss of genetic diversity. You should already be comfortable with allele frequencies and Hardy-Weinberg before tackling this.
Genetic drift is the change in allele frequencies caused by random sampling error when a finite number of individuals reproduce. It is strongest in small populations, where chance alone can push alleles to fixation (frequency = 1) or loss (frequency = 0). Drift is non-directional, reduces genetic diversity over time, and is one of the four mechanisms of evolution.
Sampling error
The difference between a sample statistic and the true population parameter, arising purely from chance when only a subset of the population is measured or reproduced. In simple terms, if you grab a handful of marbles from a jar, the colour mix in your hand probably will not perfectly match the jar.
Genetic drift
The change in allele frequencies in a population from one generation to the next due to sampling error. Think of it as evolution by luck rather than by fitness advantage.
Fixation
The point at which an allele reaches a frequency of 1.0 in a population, meaning every individual carries it. The other allele has been lost (frequency = 0). Drift can cause fixation even for neutral alleles.
Founder effect
A special case of genetic drift that occurs when a small group of individuals establishes a new population. The allele frequencies in the founding group may differ substantially from the source population by chance, and those skewed frequencies become the starting point for the new population.
Heterozygosity
The frequency of heterozygous individuals (those carrying two different alleles at a locus) in a population. Drift reduces heterozygosity over time, especially in small populations, because alleles tend toward fixation or loss.
Population size (effective)
The number of breeding individuals in a population. Smaller populations experience stronger genetic drift. Hardy-Weinberg assumes infinite population size precisely to eliminate drift.
The lecture uses a flat-tailed horned lizard population to demonstrate sampling error. A population of 32 lizards has three genotypes at a horn-length locus:
H₁H₁ (blue, small horns, 1 cm)
H₁H₂ (yellow, medium horns, 1.5 cm)
H₂H₂ (red, large horns, 2 cm)
The true population mean horn length is 1.5 cm, and the true frequency of allele H₁ is 0.5.
Students draw random samples of four lizards and estimate the population mean and allele frequency. Most students get estimates that differ from the true values. That difference is sampling error. It arises purely from chance, not from any bias in the sampling method.
When results from hundreds of students are pooled, the estimates cluster around the true value but spread out. This spread is the distribution of sampling error.
In real populations, not every individual gets to reproduce. The individuals that do reproduce are, in effect, a sample of the population. If that sample does not perfectly represent the allele frequencies of the whole population, the next generation will have slightly different allele frequencies. That change is genetic drift.
The key definition: change in allele frequencies in a population due to sampling error is called genetic drift.
Drift is stronger in small populations and weaker in large ones. This is the same principle as coin-tossing: flip a coin 10 times and you might get 7 heads; flip it 10,000 times and the proportion of heads will be very close to 0.5.
The lecture demonstrates this with simulations:
Small populations (N = ~10): genotype frequencies in the offspring generation deviate substantially from Hardy-Weinberg predictions. The "predicted vs actual" graphs look quite different.
Large populations (N = ~500-600): genotype frequencies closely match the Hardy-Weinberg predictions of 25% H₁H₁, 50% H₁H₂, 25% H₂H₂.
In computer simulations across 100 generations:
Population size 4: allele frequencies swing wildly, alleles reach fixation or loss quickly
Population size 40: moderate fluctuations, some alleles still fix over 100 generations
Population size 400: allele frequencies stay relatively stable, drifting only slightly
Smaller populations lose heterozygosity faster and fix alleles sooner.
The founder effect is a specific example of drift that occurs when a small number of individuals start a new population. The lecture uses hereditary deafness on Martha's Vineyard as a case study:
In the general US population, hereditary deafness occurs in about 1 per 5,700 people
On Martha's Vineyard, the rate was 1 in 155
In the town of Chilmark, it was 1 in 25
This happened because the founding settlers (including Jonathan Lambert and Elizabeth Eddy) happened to carry the recessive allele for deafness at a higher frequency than the general population. That initial sampling error was amplified through generations of a small, relatively isolated community.
Any change in allele frequency, including change caused by drift, counts as evolution. Drift is non-directional (it does not favour beneficial alleles) and random, but it is real evolution.
Hardy-Weinberg expected genotype frequencies (for the lizard example)
With Fr(H₁) = Fr(H₂) = 0.5: Fr(H₁H₁) = 0.25, Fr(H₁H₂) = 0.50, Fr(H₂H₂) = 0.25.
General rule for drift magnitude
No single formula is given in this lecture, but the principle is: the smaller the population, the larger the expected deviation from Hardy-Weinberg predictions due to sampling error. Drift magnitude scales inversely with population size.
The Martha's Vineyard deafness example shows how the founder effect can concentrate a rare allele in a small, isolated community. The same principle explains unusually high rates of certain genetic conditions in other founder populations, such as Tay-Sachs disease among Ashkenazi Jewish communities and Ellis-van Creveld syndrome among the Old Order Amish. Conservation biology relies heavily on understanding drift: endangered species with small population sizes lose genetic diversity through drift, which can reduce their ability to adapt to future environmental changes.
Students often think genetic drift only matters in "tiny" populations. Drift operates in all finite populations. It is simply more noticeable and consequential in small ones.
A frequent mistake is confusing drift with natural selection. Drift is random and does not favour alleles that improve fitness. Selection is directional and does.
Some students believe drift can only decrease the frequency of an allele. Drift can push an allele in either direction, up or down, because it is driven by chance.
Students sometimes think the founder effect requires the founders to be "special" in some way. The founders are just a random sample of the source population. Their allele frequencies differ from the source by chance, not by design.
⚠️ You will likely be asked to predict which simulated allele-frequency graph corresponds to a smaller vs larger population. Smaller populations show wider, more erratic swings and faster fixation.
⚠️ Know the definition of genetic drift and be able to distinguish it from selection. Drift = random, non-directional. Selection = non-random, directional.
⚠️ The founder effect is a favourite exam example. Be ready to explain the Martha's Vineyard deafness case and why allele frequencies differed from the mainland.
⚠️ Expect a question asking whether drift counts as evolution. It does, because allele frequencies change.
True or false: Genetic drift has a stronger effect in large populations than in small populations. (False. Drift is stronger in small populations.)
Fill in the blank: The change in allele frequencies due to sampling error is called ________. (Genetic drift.)
True or false: Genetic drift can cause a neutral allele to reach fixation. (True.)
Fill in the blank: When a small group of individuals founds a new population with different allele frequencies from the source, this is called the ________. (Founder effect.)
True or false: Genetic drift favours alleles that improve fitness. (False. Drift is random and non-directional.)
Q: Two simulated populations start with allele A₁ at frequency 0.5. Population X has 20 individuals, Population Y has 500 individuals. After 100 generations, which population is more likely to have lost allele A₁ entirely? Explain why.
A: Population X (the smaller one). In small populations, sampling error is larger each generation, so allele frequencies fluctuate more widely. Over many generations, these larger random swings make it more likely that an allele will drift to 0 (loss) or 1 (fixation).
Q: A small group of birds is blown to a remote island during a storm. The source population on the mainland has allele B at a frequency of 0.05. In the founding group of 10 birds, 3 carry allele B. What is the frequency of B in the new island population, and what phenomenon does this illustrate?
A: If 3 out of 10 diploid birds carry at least one copy, the exact allele frequency depends on their genotypes, but the frequency is clearly much higher than 0.05. This illustrates the founder effect: the small founding sample has allele frequencies that differ from the source population by chance.
Q: A student claims, "Genetic drift is not really evolution because it is just random noise." Is this correct?
A: No. Evolution is defined as a change in allele frequencies in a population over time. Drift produces exactly that. The mechanism is random, but the outcome (changed allele frequencies) meets the definition of evolution.
Q: Look at two simulated allele-frequency graphs over 100 generations. Graph A shows lines that fluctuate mildly and stay between 0.3 and 0.7. Graph B shows lines that swing wildly, with several reaching 0 or 1. Which graph represents a smaller population?
A: Graph B. The wider swings and more frequent fixation events are characteristic of smaller populations, where sampling error has a stronger effect.
Q: Why does genetic drift reduce heterozygosity over time?
A: Drift pushes allele frequencies toward 0 or 1 (fixation or loss). Once an allele is fixed or lost, all individuals are homozygous at that locus. The more alleles that reach fixation, the lower the overall heterozygosity in the population.
Genetic drift builds on Hardy-Weinberg equilibrium (Lectures 14-15) by violating the "infinite population size" assumption. It pairs with the mutation notes from the first half of this lecture to complete the picture of non-selective mechanisms of evolution. Later topics on speciation and conservation genetics will draw heavily on drift, particularly the founder effect and the loss of genetic diversity in small, isolated populations.
Genetic drift, sampling error, founder effect, allele fixation, loss of heterozygosity, population size and drift, small population genetics, Hardy-Weinberg violation, random evolution, non-directional evolution, Martha's Vineyard deafness, horned lizard, coin-toss analogy, population genetics, BIOL 180, Introduction to Biology, University of Washington, mechanisms of evolution, conservation genetics, bottleneck effect