Difficulty: Intermediate | Prerequisites: Allele frequency basics, Mendelian genetics
Evolution happens through four main mechanisms: natural selection (differential survival and reproduction based on heritable traits), genetic drift (random changes in allele frequency, especially in small populations), mutation (new alleles arising), and gene flow/migration (alleles moving between populations). Sexual selection is a special case of natural selection driven by mating success rather than survival. Each mechanism changes allele frequencies in different ways, and they often operate simultaneously.
Natural selection
The process by which individuals with heritable traits that improve survival or reproduction leave more offspring, causing those traits to become more common over generations. Think of it as differential reproductive success based on heritable variation.
Directional selection
Selection that favours one extreme of a trait distribution, shifting the population mean over time. Example: if larger mice survive better on Gough Island, selection pushes average size upward.
Stabilizing selection
Selection that favours intermediate trait values and acts against both extremes. This keeps the population clustered around the mean.
Genetic drift
Random changes in allele frequency caused by sampling error in small populations. Drift is strongest when populations are small and does not favour any particular allele.
Founder effect
A special case of genetic drift that occurs when a small group of individuals starts a new population. The new population's allele frequencies may differ from the source population purely by chance.
Mutation
A heritable change in DNA sequence. Mutation is the ultimate source of all new alleles and new genetic variation, but it changes allele frequencies very slowly on its own.
Gene flow (migration)
The movement of alleles between populations, typically through individuals moving and breeding. Gene flow can introduce new alleles to a population or change existing allele frequencies.
Sexual selection
A form of natural selection in which individuals with certain traits have greater mating success. It often leads to sexual dimorphism, where males and females of the same species look noticeably different.
Fitness
The reproductive success of an individual relative to others in the population. Higher fitness means more offspring that survive to reproduce.
Allele frequency
The proportion of a specific allele among all copies of that gene in a population. Think of it as how common a particular version of a gene is in the gene pool.
Heritable variation
Differences among individuals that are passed from parents to offspring through genes. Selection can only act on variation that is heritable.
Quantitative trait
A trait influenced by many genes of small effect, producing a continuous range of phenotypes (e.g. body size, height). The F2 offspring of crosses between different populations show a bell-shaped distribution of the trait.
Mutation slowly changes allele frequencies over generations.
To calculate the new frequency of allele f after one generation of mutation from f to F at rate μ: new freq(f) = freq(f) - (freq(f) × μ).
Example: freq(f) = 0.92, mutation rate = 5 × 10⁻⁴. New freq(f) = 0.92 - (0.92 × 5/10,000) = 0.92 - 0.00046 = 0.91954. Rounding: 0.9195.
Mutation alone is a very weak force. It matters because it is the only source of entirely new alleles.
Selection requires three conditions: variation in a trait, heritability of that trait, and differential fitness (some variants survive or reproduce better).
Example (steelhead trout): a gene for magnetic sense has two alleles, S (sensitive) and N (not sensitive). Genotype frequencies: 0.81 SS, 0.18 SN, 0.01 NN. If all SS return to breed, half of SN return, and no NN return:
Spawners: 81 SS (making 810 S gametes) + 9 SN (making 45 S + 45 N gametes). Total gametes: 855 S + 45 N = 900.
New allele frequencies: freq(S) = 855/900 = 0.95, freq(N) = 0.05.
Next generation genotype frequencies (assuming Hardy-Weinberg): SS = 0.95² = 0.9025, SN = 2(0.95)(0.05) = 0.10, NN = 0.05² ≈ 0.
This can be framed as selection (N allele has lower fitness because individuals carrying it are less likely to breed) or migration (N-carriers leave the breeding population at higher rates).
Drift is random fluctuation in allele frequencies due to sampling error. It is strongest in small populations.
In the mouse running experiment (Bronikowski), 8 lab populations were maintained for many generations. In 4 populations, the most enthusiastic runners were selected as breeders (selection lines). In the other 4, breeders were chosen at random (control lines).
Even in the selection lines, running distance fluctuated up and down between generations. This is drift operating on top of selection, because the populations are small.
After about 20 generations, the selection lines plateaued. This happened because strong directional selection had depleted the genetic variation for running behaviour. Without genetic variation, selection cannot produce further change.
If selection were removed (breeders chosen at random in all lines), all populations would evolve by drift alone. The selection lines would likely drift downward somewhat (since the extreme trait value is no longer being maintained) and the control lines would continue fluctuating randomly.
Gough Island mice are much larger than mainland mice. Two mechanisms could explain this:
Genetic drift / founder effect: The founding population may have had unusually high frequencies of large-size alleles by chance, or drift in subsequent small generations pushed size upward.
Natural (directional) selection: Large size may confer higher fitness in the island environment (e.g. better thermoregulation, competition, or predator defence in the absence of mainland predators).
The F2 offspring from crosses between island and mainland mice show a continuous, bell-shaped distribution of skull sizes. This tells us size is a quantitative trait controlled by many genes of small effect, and that it is heritable.
Sexual selection occurs when mating success varies among individuals, and the traits associated with mating success are heritable.
In dark-eyed juncos, males are showier (darker colours) than females. The data show a strong positive relationship between number of mates and total offspring for males, but a much flatter relationship for females.
This means males benefit more from additional matings than females do. Female reproductive success is limited more by resources than by mate availability.
Because males gain fitness from extra mates, selection favours heritable traits that increase male attractiveness (darker colour). This is sexual selection driving the evolution of showier males.
Males are not as showy as peacocks because extreme showiness carries survival costs. At some point, natural selection against being too conspicuous balances sexual selection for being attractive, resulting in stabilizing selection on male showiness.
To test whether a trait is adaptive, manipulate the trait and measure fitness.
Example: to test whether dark colour is adaptive in male juncos, randomly assign males to two groups: an unbleached control group and an experimental group whose feathers are bleached lighter. Then measure how many mates each male attracts.
Key features: random assignment to groups, manipulation of only the trait of interest, measurement of a fitness proxy (mating success).
Evolution does not happen because organisms "need" a trait. Organisms do not change in response to need.
Traits are not lost through disuse. Lamarckian inheritance (acquired traits being passed on) is incorrect.
Selection is not a "pressure" that acts on individuals to change them. Individuals do not change during their lifetimes in response to selection.
The correct explanation for the evolution of a new trait requires three elements:
There was heritable variation in the ancestral population.
Individuals with certain variants had higher reproductive success (fitness).
Over many generations, allele frequencies shifted, changing the population average.
Example (flightless birds): In the ancestral population, there was heritable variation in wing size. In an environment without predators, individuals with smaller wings may have had higher fitness (less energy spent on wings, more on reproduction). Over many generations, average wing size decreased.
Allele frequency after one generation of mutation
If allele f mutates to allele F at rate μ per generation:
f_{\text{new}} = f_{\text{old}} - (f_{\text{old}} \times \mu)Worked example: f = 0.92, μ = 5 × 10⁻⁴
f_new = 0.92 - (0.92 × 0.0005) = 0.92 - 0.00046 = 0.9195
Hardy-Weinberg genotype frequencies (used after calculating new allele frequencies)
For two alleles with frequencies p and q:
\text{freq}(AA) = p^2, \quad \text{freq}(Aa) = 2pq, \quad \text{freq}(aa) = q^2These apply when you have calculated new allele frequencies (e.g. after selection removes some genotypes from the breeding population) and need to predict the next generation's genotype frequencies, assuming random mating.
Students often say organisms evolve because they "need" to. Evolution has no foresight. Variation arises randomly, and selection filters it after the fact.
Students describe selection as a "pressure" or "force" that changes individuals. Selection does not change individuals. It changes the composition of the population by favouring some variants over others.
Students claim traits are lost through disuse (Lamarckian thinking). Traits are lost because individuals without them have higher fitness, or because mutations that degrade the trait are no longer selected against.
Students forget to specify heritable variation when explaining how selection works. Selection cannot act on traits that are not heritable.
Students confuse genetic drift with natural selection. Drift is random and does not favour any particular allele. Selection is non-random and consistently favours alleles associated with higher fitness.
Students assume a selection plateau means evolution has stopped. It means genetic variation for that trait has been depleted by selection. Other evolutionary forces (drift, mutation) still operate.
⚠️ You will be asked to calculate allele frequency changes from mutation. Memorise the formula and practise the arithmetic.
⚠️ Expect multi-part problems where you calculate allele frequencies after selection removes certain genotypes, then predict next-generation genotype frequencies using Hardy-Weinberg.
⚠️ You will need to identify which mechanism of evolution is at work in a given scenario. Be prepared to explain your reasoning (selection requires differential fitness; drift requires small population and randomness; migration requires movement between populations).
⚠️ Exam questions often present a flawed student explanation of evolution and ask you to identify the error and correct it. Know the three classic mistakes: evolution by need, loss through disuse, selection as a pressure acting on individuals.
⚠️ Sexual selection questions will ask you to interpret graphs of mating success vs. reproductive success and explain sexual dimorphism.
⚠️ You may be asked to design an experiment to test whether a trait is adaptive. Remember: random assignment, manipulation of only the focal trait, measurement of fitness.
True or False: Natural selection acts on individuals, causing them to change during their lifetimes.
False. Selection acts on populations over generations. Individuals do not change.
Fill in the blank: The only evolutionary mechanism that creates entirely new alleles is ___.
Mutation
True or False: Genetic drift is more powerful in large populations than in small ones.
False. Drift is strongest in small populations.
Fill in the blank: When males are showier than females, this is typically explained by ___ selection.
Sexual
True or False: If a selection experiment reaches a plateau, it means no further evolution of any kind can occur in that population.
False. It means genetic variation for that trait is depleted. Other traits can still evolve, and drift and mutation still operate.
Q: The frequency of allele A is 0.80 and the mutation rate from A to a is 2 × 10⁻⁴. What is the frequency of A in the next generation?
A: 0.80 - (0.80 × 2/10,000) = 0.80 - 0.00016 = 0.79984.
Q: In a population of 100 beetles, 64 are BB, 32 are Bb, and 4 are bb. Only BB and Bb beetles survive a drought. Calculate the allele frequencies in the gene pool of survivors and predict next-generation genotype frequencies.
A: Survivors: 64 BB + 32 Bb = 96. Gametes: BB contribute 128 B; Bb contribute 32 B + 32 b. Gene pool: 160 B + 32 b = 192 total. freq(B) = 160/192 = 0.833, freq(b) = 32/192 = 0.167. Next generation: BB = 0.833² = 0.694, Bb = 2(0.833)(0.167) = 0.278, bb = 0.167² = 0.028.
Q: A student writes, "Cheetahs evolved to be fast because they needed speed to catch prey." Identify the error and provide a better explanation.
A: The error is "evolution by need." Evolution does not respond to needs. A better explanation: in the ancestral population, there was heritable variation in running speed. Faster individuals caught more prey, survived better, and left more offspring. Over many generations, the average speed of the population increased.
Q: In the mouse running experiment, why did the selection lines fluctuate between generations even though selection was being applied?
A: Genetic drift. The lab populations were small, so random sampling error in which individuals bred caused allele frequencies (and therefore running behaviour) to fluctuate from generation to generation, even under directional selection.
Q: Explain why male dark-eyed juncos are showier than females, using the concept of sexual selection.
A: Males gain more fitness from additional matings than females do (the slope of offspring vs. mates is steeper for males). This means heritable traits that increase male mating success, such as darker plumage, are favoured by sexual selection. Females, whose reproductive success is limited by resources rather than mates, are not under the same selection for showiness.
Q: Why are male juncos not as showy as peacocks?
A: Extreme showiness carries survival costs (conspicuousness to predators, energy expenditure). At some point, the survival cost of being showier outweighs the mating benefit. Natural selection against extreme ornaments and sexual selection for them reach a balance, producing stabilizing selection on male showiness.
The four mechanisms of evolution connect to population genetics and Hardy-Weinberg equilibrium: when none of the mechanisms is operating (and mating is random, and the population is large), allele frequencies stay constant. Each mechanism violates one or more of the Hardy-Weinberg assumptions.
Sexual selection links to behavioural ecology and mating systems. Quantitative genetics (many genes of small effect) comes up again in agricultural breeding and human genetics.
Natural selection, genetic drift, founder effect, bottleneck effect, mutation rate, allele frequency, gene flow, migration, sexual selection, sexual dimorphism, directional selection, stabilizing selection, disruptive selection, fitness, reproductive success, heritable variation, quantitative trait, polygenic, Hardy-Weinberg equilibrium, adaptation, Lamarckism, BIOL 180, evolutionary mechanisms, selection plateau, depletion of genetic variation, experimental evolution