Biology › Section J · Evolution › Topic J.5
Population genetics and Hardy-Weinberg
Evolution, at its most measurable, is a change in the allele counts inside a gene pool. Population genetics tracks those counts, and the Hardy-Weinberg equation supplies the reference point: what a population should look like if no evolutionary force touched it. Any gap between prediction and reality becomes evidence that evolution is happening.
Genes live in populations, not just individuals
A diploid individual can carry at most two alleles for any one gene, but the population it belongs to may hold many more. Add up every allele copy in the group and you have its gene pool. Population genetics begins with a counting question: of all those copies, what fraction is one particular allele? That fraction is the allele frequency. A survey of the human ABO blood system in Jordan tallied the I-A allele at 26.1 percent of all copies, I-B at 13.4 percent, and i at 60.5 percent - three fractions that together account for the whole pool.
Evolution, redefined as a change in the numbers
When Darwin published his theory, the rules of inheritance were still a mystery - Mendel's paper appeared a few years later and went unnoticed for decades. Once genetics was rediscovered in the early twentieth century, the two fields were fused into the modern synthesis: the 1940s-era picture of how natural selection reshapes a population's genetic makeup. Out of that merger comes a sharper working definition: at the population level, evolution is a change in allele frequency from one generation to the next. Shifts on this small scale are microevolution; the grander changes that give rise to new species and higher taxonomic groups are macroevolution.
Natural selection edits the gene pool
Suppose an allele gives its carriers a survival or reproductive edge. Those carriers leave more offspring, and their offspring repeat the pattern, so the allele's share of the gene pool climbs generation by generation until it may become fixed - present in every individual. A study in Equatorial Guinea tracked pesticide-resistance alleles in mosquitoes: before an anti-malaria campaign, two resistance alleles sat at 6.3 and 74.6 percent with the susceptible allele at 19.0 percent. After two years of pesticide use, the counts had shifted to 11.1, 83.3, and 5.6 percent. The gene pool changed, which means microevolution occurred - caught in the act.
Not every shift has a reason
Frequencies can also move by accident. Genetic drift is a random change in allele frequency that gives the population no advantage - in a small group, pure luck can let one version multiply and another fade. The founder effect is a related accident: when a few individuals start a new population, whatever alleles they happened to carry become the founding stock, which may look nothing like the parent gene pool. Selection and drift usually operate side by side, so untangling which one drove a given shift is often impossible.
A baseline that never happens
In 1908, the mathematician Godfrey Hardy and the physician Wilhelm Weinberg asked what a gene pool would look like if nothing ever disturbed it. Their answer - the Hardy-Weinberg principle of equilibrium - states that allele and genotype frequencies are inherently stable: they stay constant generation after generation unless some evolutionary force is acting. The model assumes no new mutations, no migrants arriving or leaving, random mating, a population so large that drift cannot distort the odds, and no selection favoring any genotype. No natural population satisfies all five conditions - and that is exactly what makes the model useful.
The algebra of the gene pool
Give the dominant allele's frequency the name p and the recessive's the name q. If those are the only two alleles at the locus, then p + q = 1, because every copy in the pool is one or the other. What biologists usually want, though, is the genotype frequencies - the population's genetic structure. Picture drawing two alleles at random from the pool: homozygous dominant individuals turn up at a frequency of p squared, heterozygotes at 2pq, and homozygous recessive individuals at q squared. The three together account for everyone, so p squared + 2pq + q squared = 1.
Counting the ones you can see
The equation's greatest trick is turning one observable number into three hidden ones. Most genotypes cannot be read off an organism - a violet-flowered plant might be homozygous dominant or heterozygous, indistinguishable by eye. A recessive phenotype is different: it pins down the genotype exactly, because only homozygous recessive individuals show it. So count the recessives, divide by the population size to get q squared, take the square root to get q, subtract from 1 to get p - and the whole genetic structure falls out, heterozygotes included. This is how carrier rates for recessive disorders are estimated in humans.
Worked example: predicting a flower population
In a population of 400 plants, violet flowers (V) are dominant over white (v), with p = 0.6 and q = 0.4. Homozygous dominant: p squared = 0.36, so about 144 plants are VV. Heterozygous: 2pq = 0.48, so about 192 plants are Vv. Homozygous recessive: q squared = 0.16, so about 64 plants are vv. The check: 144 + 192 + 64 = 400. For color, 336 plants carry at least one V and flower violet, while the 64 vv plants flower white. Notice that the white-flowered fraction (16 percent) is far smaller than the white-allele fraction (40 percent) - recessive alleles spend most of their time hidden inside carriers.
Worked example: working backward from a rare disorder
A recessive disease turns up in about 1 of every 2,500 newborns. The affected ones are homozygous recessive, so q squared = 1/2500 = 0.0004. Take the square root: q = 0.02, and therefore p = 0.98. Now the carriers: 2pq = 2 times 0.98 times 0.02, roughly 0.039 - about 4 percent of the population, one person in twenty-five, even though only one in 2,500 shows any symptoms. A rare recessive condition thus implies a surprisingly large pool of silent carriers - and it explains why selection struggles to eliminate the allele: it cannot see the heterozygotes.
When the numbers refuse to sit still
Compare a real population against its Hardy-Weinberg prediction: any mismatch is evidence that a force is at work. Influenza makes a fast-moving case. The virus mutates and replicates so quickly that the vaccine built for last year's strains may fail against this year's - which is why flu shots are redesigned annually, with scientists tracking these frequency changes and predicting the next winners. If a population's genotype counts fit the equation, no evolutionary force is detectably acting; if they do not, the size and direction of the gap are clues about which force it is.
Common mistakes to avoid
First, treating p squared as the dominant phenotype frequency - the dominant phenotype is p squared plus 2pq, and only the recessive phenotype matches a single genotype. Second, assuming the dominant allele must be the common one - dominance is about masking, not frequency, and p can be small. Third, adding genotype frequencies to check p + q = 1 - allele frequencies sum to one, and genotype frequencies sum to one separately; never mix the two lists. Fourth, expecting a real population to sit at equilibrium - the five conditions never fully hold, and the equation is a baseline, not a description. Fifth, confusing selection with drift - selection is non-random and tracks fitness, while drift is chance and matters most in small populations.
Key takeaways
Population genetics studies evolution as change in allele frequencies within a gene pool. Selection pushes helpful alleles toward fixation, while drift and the founder effect shift frequencies by chance. The Hardy-Weinberg principle defines a non-evolving baseline - no mutation, no migration, random mating, no drift, no selection - that no real population meets, which is what makes it useful. With two alleles, p + q = 1 and p squared + 2pq + q squared = 1 predict genotype frequencies, and a recessive phenotype count gives q squared directly. Departures from the predicted values mean evolution is occurring.
Flashcards
Flashcard
The sum of all alleles in a population
Test your understanding
Question 1 of 5
If you know the allele frequencies at a two-allele locus, what does the Hardy-Weinberg equation let you calculate?
Question 2 of 5
Which of the following would break Hardy-Weinberg equilibrium?
Question 3 of 5
In a population with two alleles where p = 0.7 and q = 0.3, what fraction of individuals should be heterozygous?
Question 4 of 5
Why do recessive phenotypes give the cleanest entry point into the Hardy-Weinberg calculation?
Question 5 of 5
A viral population's allele frequencies shift markedly between one season and the next. What does this shift signify?
