Biology · Heredity, Variation & Evolution

Natural selection, genetic drift and change in populations

Begin with twenty hypothetical haploid organisms: ten carry allele X and ten Y at a focal locus. Each organism has one copy at that locus, so the initial frequency of X is 10/20 = 1/2. Now ask how the next generation is formed; different mechanisms can change the same count.

The earlier lessons on biological scales and cells provide useful background: recognise a cell, its DNA and the difference between an individual and a population. The DNA and nucleotide explanation in Chemistry of Life is optional support. You need fractions such as one half and one quarter; the worked cases explain the calculations before using them. Detailed molecular copying, Mendelian inheritance and evolutionary inference return in later chapters.

Back to the Biology map

By the end, you can

Calculate an allele-frequency change, distinguish the assumptions of selection and drift models, and interpret fitness in reproductive context.

Measure inherited change across generations

Allele frequency is the fraction of relevant allele copies represented by a particular version. In a diploid population, count copies at the locus rather than treating each individual as one allele; an Aa individual contributes one A and one a. Our first population is haploid to keep this accounting visible.

Evolution includes inherited changes in populations across generations, such as changes in allele frequency. The organisms that contribute to the next generation can differ in survival and reproduction, but the population's frequency is a collective record.

An individual becoming larger during development is a different event. Even a change in a population's average phenotype does not alone establish an inherited evolutionary change, because conditions and developmental composition may have changed.

Connect inheritance to reproductive contribution

Natural selection occurs when heritable differences are associated with differences in reproductive contribution. Survival can affect that contribution, but survival alone is not the whole record. A variant's consequences depend on the environment and on the rest of the organism's biology.

In the supplied clonal model, every X-bearing parent contributes two offspring reaching the counted next-generation stage; every Y-bearing parent contributes one. Ten X parents therefore contribute twenty X offspring, and ten Y parents contribute ten Y offspring. X rises from 10/20 to 20/30, or two thirds.

Fitness concerns reproductive contribution in a specified context. It is not a synonym for bodily strength, size, health or a moral judgement. The model assigns X twice Y's contribution in this setting; it does not make X universally superior or predict that X must take over in every environment.

The model holds mutation, migration and random sampling aside to make the selection relationship visible. Real populations can experience these processes together. Selection also acts on existing variation; it does not cause organisms to request the allele they need.

Reproductive contribution changes the counted mixture

Deterministic clonal model, offspring counted at the specified stage. No mutation, migration or random sampling. Token count represents organisms; frequency denominators change with population size.

Read the diagram in words

X contributes twenty of thirty offspring, so its frequency rises from one half to two thirds. Y still contributes ten organisms, while the total changes.

Parent generation
Allele	Count
X	10
Y	10
Total	20

X: 10/20 = 1/2

Counted next generation
Allele	Count
X	20
Y	10
Total	30

X: 20/30 = 2/3

Each X parent contributes two counted offspring
Each Y parent contributes one counted offspring

One haploid organism with one allele copy at the focal locus

Chance can change the mixture without a fitness difference

In the neutral drift model, X and Y have equal chances of being copied into each offspring. Each of ten new offspring copies one uniformly chosen parent's allele from the twenty-parent pool, with replacement. X therefore has probability 1/2 on every draw, but the realised ten-copy mixture can differ from half.

One possible outcome contains eight X and two Y copies, giving frequency 8/10 = 0.8. Another contains three X and seven Y, giving 3/10 = 0.3. These invented outcomes show directions that are possible under equal reproductive chances; they are not actual random trials.

Genetic drift is change due to chance sampling in finite populations. It tends to have stronger proportional effects in smaller populations and does not systematically supply useful adaptations. A changed frequency by itself does not tell us whether selection, drift, gene flow or another process caused it.

The ten-copy generation here is deliberately smaller than the twenty-copy parent pool; it is an illustrative sampling model. A real reduction in population size need not be neutral, and chance and selection may both affect which hereditary copies remain.

Equal chances allow different realised outcomes

Invented possible outcomes, not observed or generated random trials. The outcomes are alternatives, not successive generations. Sampling size changes from twenty parents to ten offspring; no differential fitness is assigned.

Read the diagram in words

Each outcome begins from the same half-X parent pool. One possible result is 0.8 X; another is 0.3 X. Equal per-draw chances do not guarantee a half-X realised generation.

Equal chances allow different realised outcomes
Pool or alternative outcome	X copies	Y copies	Total	Fraction of X copies
Same parent pool	10	10	20	10/20 = 0.5
Possible outcome 1	8	2	10	8/10 = 0.8
Possible outcome 2	3	7	10	3/10 = 0.3

Ten independent uniform draws, with replacement
From	To	Relationship
Same parent pool	Possible outcome 1	Ten independent uniform draws, with replacement
Same parent pool	Possible outcome 2	Ten independent uniform draws, with replacement

Chance of X on each draw: 1/2

Worked example

Use the denominator that belongs to the generation

Selection gives twenty X plus ten Y offspring, for thirty in all. Dividing X by the old population size would give 20/20 = 1 and the wrong frequency. Use the new total: 20/30 = 2/3. Compare with the old 10/20 = 1/2; the increase is 1/6, about 16.7 percentage points. This is an exact property of the constructed deterministic case, not a measured effect or an estimate of a mutation rate.

Check your understanding

Why is X's next-generation frequency two thirds rather than one?
Show reasoning
The next generation contains thirty organisms: twenty X and ten Y. The relevant fraction is 20/30.
Can a chance-driven frequency increase demonstrate that X has higher fitness?
Show reasoning
No. The drift model allows an increase even with equal reproductive chances. Fitness differences need separate evidence.
Would one parent's leaf growing longer constitute population evolution?
Show reasoning
No. That is an individual phenotypic change. An evolutionary claim concerns inherited changes across population generations.

What to carry forward

Count inherited variants, explain the reproductive process and keep mechanism claims bounded.

References

Resources supporting this lesson’s explanations and examples.