Shared article

Genetic Variation in Evolution

Heritability, selection and drift

The physical differences between organisms of the same species are called phenotypic variation. Phenotypic variation is what natural selection acts on – one individual has a greater fitness than another because it is stronger, faster, more camouflaged, or just more attractive to the opposite sex. As luck plays an important role, a fitter individual will not always survive and produce more offspring than a less fit one. However, averaged over many generations some traits will tend to be passed on while others are lost. Phenotypic variation is caused by both environmental factors such as temperature or nutrient availability, and genetic variation – differences in the DNA sequence between individuals.

This variation can be represented in the following equation:

Phenotype = Genotype + Environment


P = G + E

Genotypic effects on variation can be divided into additive effects, dominant effects and interaction effects. These three factors together are referred to as broad-sense heritability.

As the dominant and interaction effects are due to the particular combination of genes in an individual, and genes are shuffled during meiosis, only the additive effects are actually passed on from parent to offspring. This is referred to as narrow-sense heritability and is what is usually considered in studies into the effects of genetic variation on evolution.

  • Broad-sense heritability = genetic effects on organismal phenotype
  • Narrow-sense heritability = genetic effects that are directly passed on from parent to offspring

Selection can only act on the additive genetic variation as this is the component of variation that is passed on to the next generation. The population geneticist Ronald Fisher showed that the response to selection is directly proportional to the additive genetic variation in the population. This is known as Fisher’s Fundamental Theorem.

Two major forces act to change the frequency of alleles in a population: selection and drift. Selection changes allelic frequency in a directional manner and has a greater effect in larger populations with greater additive genetic variation. Genetic drift is essentially a random process, analogous in some ways to Brownian motion, where allelic frequencies increase or decrease randomly until the allele is lost completely or becomes fixed in the population. Genetic drift can cause the loss of advantageous alleles or the fixation of deleterious ones provided that the population is small enough and the selective pressure is weak enough. This has consequences for conservation work. The small surviving populations of highly endangered species often show reduced fitness due to the fixation of deleterious alleles and reduced genetic variation. This can lead, for example, to increased susceptibility to disease.

Genetic variation can arise in several ways, for example through gene duplication, inversions, and deletions. However, here we will focus on single-nucleotide polymorphisms (SNPs, pronounced as “snips”) caused by point mutations of single DNA bases.

Negative/purifying selection

Negative or purifying selection is what results from a mutation that reduces the fitness of the individuals that carry it. It is easy to see why negative selection is more common than positive selection. If you think of a protein as a machine, then making a random change to one of the parts of that machine will more likely reduce or destroy its function than improve it. Therefore natural selection will act to remove the mutated allele from the population.

Positive selection

Much more rarely a mutation in a gene will produce an allele that confers a higher fitness on the individuals that carry it. In this case natural selection will act to increase the frequency of the mutated allele in the population. In theory this will lead over time to the fixation of the new allele in the population and the loss of the old one.

Exceptions – dominance, timing effects and polymorphisms

In nature the situation is rarely as simple as described above. As with most situations the exceptions are more interesting than the rules.


Sexually reproducing organisms have two copies of each chromosome, one from each parent. The sister chromosomes carry the same genes. However, the particular alleles on the sister chromosomes can be different. This can often lead to dominance effects, where the effect of one allele can dominate the phenotype to varying degrees.

The dominance of an allele affects how it is acted on by natural selection and how it is involved in evolution. A very dominant allele will always be visible to natural selection whenever it is carried by an organism, but a recessive one will only be expressed when it 2 copies of it are carried. This means that the selective force on a recessive gene is relatively less.

An example of how the dominance of a gene can be relevant to evolution can be seen in the genetic disease cystic fibrosis. Only individuals that carry 2 copies of the cystic fibrosis allele have the disease. This means that natural selection can only act on the allele when 2 copies of the allele are present. If only 1 copy is present then the allele is silent. This means that so long as disease-causing recessive alleles are rare in the population they can persist by “flying under the radar” of natural selection.

The maintenance of recessive alleles in this way is why we see inbreeding depression in small populations. Offspring are more homozygous than the ancestral population and so show more deleterious effects of recessive alleles.

The frequency of recessive alleles in the population can be measured through sampling. In the case of the allele for sickle cell disease this frequency is much higher in some areas of Africa than would be expected based on the high fitness cost of homozygosity. This is because the heterozygous genotype confers increased resistance to certain kinds of malarial infection. This is an example of heterozygote advantage, also known as overdominance, where the heterozygote has higher fitness than either homozygote in a specific environment.

Blood smear - sickle cell

Heterozygosity can become fixed through gene duplication. One allele is duplicated within the same chromosome and then one of the copies recombines with the other allele to produce a single chromosome carrying both alleles.

fixation of heterozygosity

Expression timing (fitness vs health)

Deleterious dominant mutations are fully visible to natural selection, yet some are still maintained in the population. An example is Huntington’s disease, which can occur through spontaneous extension of CAG repeats, but is often inherited. The important point here is that the effects of Huntington’s very often do not become visible until the individual is old enough to have had children. As such, despite being a debilitating and terminal disease, in an evolutionary sense the fitness of the individual is not greatly affected. This highlights the important difference between ‘fitness’ and ‘health’ – natural selection doesn’t mind if you decay and die young, so long as you have lots of children first!


So far we’ve been working under the assumption that there is one particular allele (or one particular pair of alleles in the case of heterozygote advantage) that is the best possible in a particular situation. However, it is important to appreciate that organisms are in competition with others of the same species (conspecifics) to avoid predation and acquire resources and mates. Competing with conspecifics is like playing a card game – the best strategy often depends on what everyone else is doing. This can lead to polymorphisms within a population.

A polymorphism is when two or more genetically distinct phenotypes exist within a population. A polymorphism is maintained by negative frequency-dependent selection. This is when the fitness value of an allele decreases as that allele becomes more frequent. This prevents the allele from reaching fixation in a population. This can be illustrated using prey switching. For example the grove snail Cepaea nemoralis shows polymorphisms in the pattern and colour of its shell. The thrush predators of these snails learn to identify the shells by appearance. The snails of the rarest morph will have the highest fitness because the birds will be less able to recognise and eat them! This frequency-dependent selection by predators is called apostatic selection.

C nemoralis

Cyclic polymorphisms – side-blotched lizards

Male side-blotched lizards

In small populations of seasonally breeding animals it is possible to see cyclic changes in allele frequency. An example is the side-blotched lizards of the species Uta stansburiana. In these lizards the males exist in 3 genetically-determined morphs, each with a different throat colour and mating strategy:

  • Orange – large territory with many females
  • Yellow – sneaky males, steal females from orange males
  • Blue – small territory with one female, can fend off sneaky males

u stansburiana male morphs

cycle of fitness

As illustrated in the above figure, the fitness of each morph of male depends on the frequency of the morph over which it has an advantage. Blue males do best when the number of yellows is high, orange when blue is high, and yellows when orange is high, much like the game Rock, paper, scissors. As a result the relative frequencies of these morphs rise and fall in turn with a frequency of around 4-5 years.

male frequency oscillation

This is an idealised graph to show the oscillations of the 3 morphs over time. In reality the shape of the curves, the displacement of the curves relative to each other and the amplitude of the oscillations would not be equal between morphs. It is important to remember that averaged over time the fitness of the 3 morphs is equal, but the absolute number of individuals doesn’t have to be.

Female side-blotched lizards

In addition, the females exist as two morphs with different breeding strategies:

Orange – many small eggs, quantity

Yellow – few large eggs, quality

female frequency

In years with high population density more yellow females survive, as they hatch out larger than orange ones and can outcompete them. However this means that in the following year there are fewer females, less competition for resources, and the orange females do better. This 2-year cycle reinforces itself as each morph becomes better adapted to their respective strategies of producing higher quantity or greater quality of offspring.

An important point with both the female and male cycles is that the morph frequency cycle lags behind the morph fitness cycle. For example the yellow female fitness is highest when the orange female frequency is highest. It is this lag that gives rise to the cyclic changes.

Polymorphism can lead to speciation if the different morphs become reproductively isolated, which could happen for example if they are suited to different habitats.

The Molecular Clock

Often SNPs in DNA have no effect on evolution at all. The 4 nucleotides in DNA, adenine, cytosine, guanine and thymine code for the amino acids in proteins in non-overlapping sets of 3. This means that there are 64 possible codon sequences. Only 22 amino acids are used to make proteins. This means that the genetic code is degenerate as multiple different codons can produce the same amino acid.

What this means in practice is that the majority of mutations in the coding regions of DNA are silent; they produce no change in the sequence or structure of the protein product and are referred to as synonymous mutations. As the phenotype is not altered, natural selection does not act on these neutral changes. This means that these silent point mutations should in theory accumulate at a constant rate within the genome. This is known as the Molecular Clock. If it is assumed that the rate of synonymous mutations is relatively constant, the divergence in the nucleotide identities at synonymous sites can be used to estimate how long ago the species diverged from one another. In practice the picture is more complex; the clock must be calibrated using known divergence dates from the fossil record and in many cases the clock is not constant in time or between species.

Tests for selection

The accumulation of synonymous mutationscan be compared to the accumulation of non-synonymous mutations to infer the past evolution of a DNA sequence or gene. These comparisons are made between the species of interest and a sister group containing multiple species.

Therefore the value of dN/dS can give an indication of recent selective pressures. The assumption is that the rate of change of synonymous sites is unaffected by natural selection, while the rate of change of non-synonymous sites varies.

  • If dN/dS > 1, this implies positiveselection
  • If dN/dS < 1, this implies negativeselection
  • If dN/dS = 1, this implies no selection


Pseudogenes are gene that have been released from selective forces and have acquired synonymous and non-synonymous mutations to the extent that the gene product is no longer functional. Instinctively it seems that a gene that is not subject to selection should simply stay the same. However, as the DNA is replicated over and over again the information it encodes will decay if there is no selective process to maintain it – “Use it or lose it”. It is like a photocopy of a photocopy: unless there is some selection for the photocopy to accurately reflect the original then the information in the document will be gradually lost until it is so illegible as to be useless.

For example in both baleen whales and birds we see a pseudogene form of the Enamelin gene. Enamelin is essential for the formation of enamel in teeth and not surprisingly we see a loss of teeth in both of these unrelated groups as adaptations to filter-feeding and keratin beak feeding respectively.It is important to remember that synonymous mutations only become fixed in a population by the random action of drift. Until then they exist as SNPs. These SNPs are what are examined in paternity tests and DNA comparisons in criminal investigation. There are tens of millions of SNPs in the human genome.


Genetic variation is the raw material on which natural selection acts and is generated by mutation and imperfect DNA replication. Fitter individuals reproduce at the expense of less fit ones, but fitness is a relative concept that changes in time and space. Fitness is also dependent on the behaviour of other individuals of the same species. This can lead to polymorphic traits, maintained by negative frequency-dependent selection. If there is no selective pressure then genetic information is lost over time.

Further reading

Bleay C, Comendant T, Sinervo B (2007) An experimental test of frequency-dependent selection on male mating strategy in the field. Proc. R. Soc. 274, 2019-2025

Sinervo B, Svensson E, Comendant T (2000) Density cycles and an offspring quantity and quality game driven by natural selection. Nature 406, 985-988

Fisher R (1930) The Genetical Theory of Natural Selection. Clarendon

Al-Hashimi N et al (2010) The Enamelin Genes in Lizard, Crocodile, and Frog and the Pseudogene in the Chicken Provide New Insights on Enamelin Evolution in Tetrapods. Mol Biol Evol 27(9): 2078-2094

Deméré T A et al (2008) Morphological and molecular evidence for a stepwise evolutionary transition from teeth to baleen in mysticete whales. Syst Biol 57(1): 15-37

Williams N et al (2005) Sickle Cell Trait and the Risk of Plasmodium falciparum Malaria and Other Childhood Diseases. J Infect Dis 192 (1): 178-186

Photograph attributions

"Sickle cell blood smear" by Racheal A of

"Variation" showing snail polymorphisms by duckinwales of

Uta stansburiana photos:

"orange-blue uta" by randomtruth of  (left)

"UTST" by Bryan Hamilton of  (centre)

"another light blue uta" by randomtruth of (right)


Fastbleep © 2019.