- Selection Intensity is a measure of how strongly individuals are selected from a population to become parents of the next generation. In quantitative genetics, selection intensity describes the standardized strength of selection and helps determine how strongly a population is expected to respond when individuals with particular phenotypes or genetic values are chosen for reproduction.
- Selection intensity is commonly represented by the symbol i and is defined as the selection differential divided by the phenotypic standard deviation:
- i = S / σP
- where S is the selection differential and σP is the phenotypic standard deviation of the trait. Because both the selection differential and phenotypic standard deviation are measured in the same units, selection intensity is dimensionless.
- The selection differential is the difference between the mean phenotype of selected individuals and the mean phenotype of the original population. Selection intensity standardizes this difference relative to the amount of phenotypic variation in the population. This allows the strength of selection to be compared across traits or populations with different measurement scales.
- For example, suppose the average height of a plant population is 100 cm and the plants selected for breeding have an average height of 110 cm. If the phenotypic standard deviation is 20 cm, the selection differential is 10 cm and selection intensity is:
- i = 10 / 20 = 0.5
- The value of 0.5 indicates that the selected group is, on average, half a phenotypic standard deviation above the population mean.
- Selection intensity is strongly influenced by the proportion of individuals selected. If only a small percentage of individuals are selected as parents, the selection intensity is generally high because the selected individuals are concentrated toward the extreme end of the phenotypic distribution. If a large proportion of individuals is selected, selection intensity is lower because the selected group is closer to the population mean.
- For a normally distributed trait under directional truncation selection, the relationship between the proportion selected and selection intensity can be calculated from the normal distribution. Selecting approximately half of the population produces a relatively modest selection intensity, whereas selecting only a small fraction produces a much larger intensity. The exact value depends on the proportion selected and the assumptions of the selection model.
- Selection intensity should not be confused with selection differential. Selection differential is expressed in the original units of the trait, such as kilograms, centimeters, liters, or days. Selection intensity is standardized and dimensionless. The two are related by:
- S = iσP
- This equation shows that the same selection intensity can produce different selection differentials in populations with different amounts of phenotypic variation. A population with greater phenotypic standard deviation will have a larger selection differential for the same selection intensity.
- Selection intensity is also directly connected to the Breeder’s Equation. The basic prediction of selection response is:
- R = h²S
- where R is selection response and h² is narrow-sense heritability. Substituting the relationship between selection differential and selection intensity gives:
- R = i h² σP
- Because narrow-sense heritability is the ratio of additive genetic variance to phenotypic variance, this can also be expressed as:
- R = iσA
- where σA is the additive genetic standard deviation. This form highlights a fundamental principle of quantitative genetics: expected response depends on selection intensity and the amount of additive genetic variation available in the population.
- Selection intensity therefore influences how rapidly a population can change under artificial selection. Increasing selection intensity generally increases the expected response per generation, provided that the assumptions of the selection model remain appropriate and sufficient genetic variation is available.
- However, greater selection intensity does not always produce greater long-term genetic improvement. Extremely intense selection can reduce the number of individuals contributing genes to the next generation. This can reduce effective population size, increase inbreeding, increase the loss of genetic diversity, and potentially reduce the population’s ability to respond to future selection.
- The proportion of individuals selected is therefore an important component of breeding-program design. A breeder selecting only a small fraction of candidates may achieve high selection intensity, but this approach can increase the risk of inbreeding or loss of favorable genetic diversity. A less intense selection strategy may provide more sustainable genetic improvement over many generations.
- Selection intensity also differs between males and females in many breeding systems. If different numbers of males and females are selected as parents, the selection intensity applied to each sex may differ. Because the sexes can contribute differently to the next generation, breeding programs often consider sex-specific selection intensity when predicting genetic gain.
- For example, a breeding program may select a very small proportion of males because each male can produce many offspring, while selecting a larger proportion of females. The resulting selection intensities can therefore differ between the two sexes, contributing differently to the overall rate of genetic improvement.
- Selection intensity is also influenced by the generation interval. Selection intensity determines how strong selection is within a generation, while generation interval determines how quickly generations are replaced. A breeding program may increase annual genetic gain not only by increasing selection intensity but also by reducing the time required to produce the next generation.
- This relationship is particularly important in genomic selection. Genome-wide genetic information can allow breeders to identify promising individuals at a young age. By reducing the generation interval while maintaining or improving selection accuracy, genomic selection can substantially increase the rate of genetic gain even when selection intensity itself does not change dramatically.
- Selection intensity is therefore only one component of the overall rate of genetic improvement. A commonly used conceptual relationship for annual genetic gain includes selection intensity, selection accuracy, additive genetic standard deviation, and generation interval:
- ΔG/year ≈ (i × r × σA) / L
- where i is selection intensity, r is selection accuracy, σA is additive genetic standard deviation, and L is generation interval.
- This relationship shows that genetic improvement can be increased through stronger selection, more accurate identification of genetically superior individuals, greater additive genetic variation, or shorter generation intervals. In practice, these factors must be balanced against population health, genetic diversity, economic costs, and breeding objectives.
- Selection intensity is particularly important when selection is based on phenotype. The phenotypic distribution determines how far selected individuals lie from the population mean. Environmental variation can influence this distribution, meaning that high phenotypic selection intensity does not necessarily correspond to high genetic selection intensity.
- For traits with low heritability, phenotype may be a relatively poor indicator of genetic merit. Individuals with extreme phenotypes may have achieved those phenotypes partly because of favorable environmental conditions. Selecting them can produce a large selection differential but a smaller-than-expected genetic response.
- For this reason, breeding programs often use breeding values or genomic estimated breeding values (GEBVs) rather than relying solely on observed phenotypes. Selection based on predicted genetic merit can improve selection accuracy and make the relationship between selection intensity and genetic response more predictable.
- Selection intensity also interacts with genetic architecture. For traits controlled by many genes, selection changes allele frequencies across numerous loci. The amount and distribution of additive genetic variation determine how much response remains possible as selection continues.
- As favorable genetic variants become more common, additive genetic variance may decline. This can reduce future selection response even if selection intensity remains high. Therefore, maintaining genetic diversity is important for long-term breeding progress.
- Selection intensity can also affect multiple traits simultaneously. When selection is applied to one trait, correlated changes may occur in other traits because of genetic covariance and genetic correlation. A very strong selection intensity for one trait can therefore produce undesirable correlated responses in another trait.
- In multiple-trait breeding programs, selection indices can be used to combine several traits into an overall selection criterion. The effective selection intensity then applies to the index rather than necessarily to each individual trait separately. This allows breeders to balance competing objectives and reduce unfavorable correlated responses.
- Selection intensity is also relevant to natural selection. In natural populations, individuals with different phenotypes may have different probabilities of survival or reproduction. Strong differences in reproductive success can create strong selection, although the precise relationship between fitness and selection intensity depends on the population and the traits involved.
- In evolutionary biology, selection intensity can vary across environments and generations. Environmental changes, competition, predation, disease, climate, and resource availability can alter which phenotypes are favored. Consequently, selection intensity is not necessarily constant over time.
- The concept also interacts with genotype–environment interaction (G×E). A genotype that performs well in one environment may not have the same advantage in another. Selection intensity applied in one environment may therefore produce a different genetic response when the population is moved to a different environment.
- Selection intensity should also be distinguished from selection differential and selection response. Selection intensity describes the standardized strength of selection. Selection differential describes the phenotypic difference between selected individuals and the original population. Selection response describes the change in the population mean in the subsequent generation. These three concepts describe different stages of the selection process.
- The basic relationship can be summarized as:
- Selection intensity → Selection differential → Genetic response
- More precisely, selection intensity determines the standardized strength of selection, selection differential describes the resulting phenotypic difference between selected individuals and the original population, and additive genetic variation determines how much of that difference is transmitted to the next generation.
- Selection intensity is also important in experimental evolution and artificial selection experiments. Researchers can deliberately select individuals from one extreme of a population distribution and monitor changes across generations. Comparing selection intensity with observed response provides information about heritability, genetic variance, and the capacity of a population to evolve.
- In conservation genetics, selection intensity must be considered alongside population size and genetic diversity. Extremely strong directional selection in a small population may increase genetic drift and inbreeding while reducing the diversity required for future adaptation. Conservation strategies therefore often prioritize maintaining sufficient effective population size and genetic variation rather than maximizing short-term selection intensity.
- Selection intensity can also differ among breeding objectives. For traits that are inexpensive and easy to measure, breeders may evaluate many individuals and apply relatively strong selection. For expensive or difficult-to-measure traits, the number of candidates that can be evaluated may be smaller, limiting practical selection intensity.
- Modern breeding technologies can change these constraints. Genomic selection, high-throughput phenotyping, automated measurement, and reproductive technologies can increase the number of candidates evaluated and improve the accuracy of selection decisions. These advances can allow breeders to achieve greater genetic gain without necessarily relying on extremely high selection intensity.
- Selection intensity therefore represents one of the most important levers available for changing the rate of genetic improvement. However, its effects cannot be considered independently of heritability, additive genetic variance, selection accuracy, generation interval, genetic correlations, population size, and breeding objectives.
- Understanding selection intensity provides a foundation for understanding selection differential, selection response, the Breeder’s Equation, breeding value, genetic gain, and genomic selection. Together, these concepts explain how the strength of selection influences genetic change and why effective breeding programs must balance short-term selection pressure with long-term genetic sustainability.
