Natural selection changes allele frequencies when some heritable variants leave more offspring than others. Hardy–Weinberg equilibrium models the opposite situation, letting biologists test whether selection (or other forces) are acting in real populations.

Why “no natural selection” matters in Hardy–Weinberg

In the Hardy–Weinberg (H–W) model, no natural selection means all genotypes have equal fitness, so differences in survival or reproduction do not systematically favor any allele. If this condition holds (along with the other H–W assumptions), allele frequencies remain constant across generations.

Punnett-square visualization of Hardy–Weinberg expectations for a two-allele locus. The square partitions genotype frequencies into $p^2$ (homozygous dominant), $2pq$ (heterozygous), and $q^2$ (homozygous recessive), showing how random union of gametes produces a stable genotype distribution when Hardy–Weinberg conditions are met. Source

Selection violates H–W because it creates consistent differences in reproductive output among genotypes. Even small fitness differences can produce detectable shifts in genotype frequencies over time.

What “no selection” looks like biologically

A population approximates “no selection” when:

• Environmental conditions do not make any genotype consistently more likely to survive to reproduction.

• Mating and fertilisation do not systematically favour one genotype’s gametes (no genotype-linked fertility differences).

• Any genotype differences affect traits that are neutral with respect to survival and reproduction in that environment.

Conversely, selection is likely when genotype is associated with:

• Viability differences (some die more before reproducing)

• Fertility differences (some produce fewer gametes or offspring)

• Sexual selection (mate choice changes reproductive success)

Hardy–Weinberg equilibrium as a null hypothesis

H–W equilibrium provides a baseline expectation for genotype patterns in the absence of selection.

Worked example figure illustrating how measured allele frequencies ($p$ and $q$) are converted into predicted Hardy–Weinberg genotype frequencies. The diagram connects allele counting in a population sample to the expected genotype proportions $p^2$, $2pq$, and $q^2$, emphasizing why Hardy–Weinberg functions as a baseline (null) expectation for comparison to real data. Source

This is why the syllabus states that no natural selection occurs, making Hardy–Weinberg equilibrium a valuable null hypothesis for studying evolution.

Scientists use H–W as a null hypothesis: if observations match the H–W expectation, there is no evidence (from that dataset) that selection is operating on the locus being studied.

A key idea for AP Biology: failing to reject the null hypothesis does not prove selection is absent; it means the data do not provide sufficient evidence of selection at that time, at that locus, given sampling limits.

How “no selection” is tested using Hardy–Weinberg logic

When applying H–W as a null model, biologists typically:

• Measure observed genotype frequencies in a population sample.

• Use the H–W equilibrium expectation as the predicted genotype distribution under no selection.

• Compare observed vs expected values using a statistical test (often chi-square), interpreting discrepancies as evidence that at least one H–W condition is violated.

For this subsubtopic, the interpretation focus is:

• If genotype frequencies differ from the no-selection expectation in a way consistent with differential survival/reproduction, this supports selection as a plausible cause.

• If the difference is not statistically significant, the data are consistent with no selection (and with other assumptions not being detectably violated).

What can mimic selection signals

A mismatch with the H–W expectation does not uniquely identify selection, because other processes can also disrupt equilibrium patterns. Therefore, when the null hypothesis is rejected, biologists consider alternative explanations such as:

• Non-random mating (changes genotype frequencies without necessarily changing allele frequencies immediately)

• Migration or population structure (sampling a mixture of subpopulations)

• Mutation (usually small per generation, but relevant in some systems)

• Genetic drift in small populations (random deviations)

For AP-level interpretation, the crucial skill is recognising that H–W is a baseline: rejecting it means “some evolutionary condition is violated,” and additional evidence is needed before attributing the deviation specifically to selection.

Why this null model is powerful for studying evolution

Using “no natural selection” as a baseline helps biologists:

• Detect whether evolution is occurring at a particular gene by checking for non-equilibrium genotype patterns

• Compare populations or time points to see whether changing environments are associated with departures from the no-selection expectation

• Identify candidate loci potentially under selection for deeper investigation (e.g., linking genotype to survival or reproductive output)

In short, the “no natural selection” assumption is less a claim about reality and more a controlled reference point that makes evolutionary change measurable.