- Correlated Response to Selection is the change in the genetic mean of one trait that occurs when selection is applied primarily to another genetically correlated trait. In animal breeding, selection rarely affects only the trait directly included in the selection criterion because economically important and biologically related traits often share genes. As a result, selection for one trait can produce a favorable, unfavorable, or negligible genetic change in another trait. Correlated response is therefore an important concept for understanding genetic response, genetic correlation, selection objectives, and the consequences of multiple-trait selection.
- The genetic basis of correlated response is the genetic covariance between traits, particularly the covariance between their additive genetic values. If two traits are controlled partly by the same genes, selection that changes the frequency of alleles affecting one trait can also change the genetic level of the other trait. The direction and magnitude of this indirect change depend on the genetic correlation, the amount of additive genetic variation, the selection intensity, the accuracy of selection, and the relationship between the selection criterion and the breeding objective.
- The basic components of a phenotype can be represented as P = G + E, where P is phenotypic value, G is genetic value, and E is environmental influence. For correlated response, the important component is the additive genetic value, because additive effects are transmitted from parents to offspring in a predictable way. Dominance effects, epistatic effects, and environmental effects may influence the observed phenotypes and relationships between traits, but the predictable response to selection is primarily determined by additive genetic variation and covariance.
- Genetic correlation is central to correlated response. It measures the association between the additive genetic values of two traits and can be expressed as r_A = Cov_A(X,Y) / (σ_A,X × σ_A,Y). A positive genetic correlation means that genes increasing one trait tend, on average, to increase the other trait. A negative genetic correlation means that genes increasing one trait tend to decrease the other. A genetic correlation close to zero indicates little linear additive-genetic association, although other biological or environmental relationships may still exist.
- For example, suppose selection is applied to increase growth rate in beef cattle and growth rate has a positive genetic correlation with mature body weight. Selecting animals with high genetic merit for growth may therefore cause mature body weight to increase even if mature body weight was not directly included in the selection criterion. This is a correlated response. Such a response may be desirable when increased mature size improves production efficiency, but undesirable if it increases maintenance requirements or creates reproductive, health, or management problems.
- The correlated response can be expressed using a simplified quantitative-genetic relationship. When selection is directly practiced on trait X, the expected correlated response in trait Y is approximately CR_Y = i × r_X × (Cov_A(X,Y) / σ_P,X), where CR_Y is the correlated response in trait Y, i is selection intensity, r_X is the accuracy of selection for trait X, Cov_A(X,Y) is the additive genetic covariance between traits X and Y, and σ_P,X is the phenotypic standard deviation of the selection trait. An alternative form using genetic correlation is CR_Y = i × r_X × r_A × σ_A,Y, when the relevant assumptions are satisfied. These relationships show that stronger genetic association and greater additive genetic variation in the correlated trait can increase the magnitude of indirect response.
- Correlated response should be distinguished from correlated phenotypic change. Two traits may be phenotypically correlated because of shared environmental conditions, management, measurement effects, or genetic factors. Selection response, however, depends primarily on the genetic relationship between traits. Therefore, a strong phenotypic correlation does not necessarily mean that selection for one trait will produce a strong correlated genetic response in another trait.
- The magnitude of correlated response depends strongly on heritability. Heritability can be expressed as h² = σ²_A / σ²_P, where σ²_A is additive genetic variance and σ²_P is phenotypic variance. When the selected trait has low heritability, phenotype-based selection may be relatively inaccurate, reducing the direct and correlated responses. Information from relatives, progeny testing, repeated records, indicator traits, pedigree information, and genomic information can improve the accuracy of selection and consequently increase the expected genetic response.
- The genetic covariance between traits is particularly important because it determines the direction of correlated change. Positive covariance generally produces a positive correlated response, whereas negative covariance can produce an unfavorable response. For example, selection for increased milk production may be genetically associated with changes in fertility or health traits. If the genetic relationship is unfavorable, strong selection for production alone may unintentionally reduce reproductive performance or health.
- This is one reason why modern animal breeding programs rarely focus exclusively on a single trait. Instead, breeders often use multiple-trait selection, selection indexes, or genetic evaluation models that combine information from several economically and biologically important traits. A selection index can be represented as I = b₁x₁ + b₂x₂ + … + bₙxₙ, where x values represent information used for selection and b values are index weights. By combining traits appropriately, breeders can increase desirable responses while limiting unfavorable correlated responses.
- The breeding objective also plays an important role. A breeding objective can be represented as H = a₁A₁ + a₂A₂ + … + aₙAₙ, where A represents additive genetic merit for different traits and a represents their economic or strategic importance. The purpose of a selection index is to use available information to predict the overall breeding objective rather than maximizing a single trait in isolation. This allows correlated genetic relationships among traits to be incorporated into breeding decisions.
- Correlated responses are particularly important for production traits, reproductive traits, health traits, feed efficiency, survival, longevity, and adaptation traits. Selection for rapid growth, for example, can affect mature size, feed requirements, carcass composition, reproductive performance, or metabolic health. Selection for milk production can be associated with fertility, udder health, longevity, and energy balance. Selection for litter size can influence survival, maternal ability, and reproductive performance. These relationships demonstrate why genetic improvement should be evaluated across the complete breeding objective.
- In dairy cattle, selection for higher milk yield may produce correlated changes in milk composition, fertility, health, body condition, and longevity when these traits have genetic associations. In pigs, selection for growth rate or lean meat percentage can be associated with changes in feed efficiency, reproductive traits, carcass quality, or meat quality. In poultry, selection for egg production may influence egg quality, persistency, body weight, and reproductive traits. In sheep, selection for growth, wool production, or litter size may produce correlated changes in mature size, reproductive performance, maternal ability, or survival.
- Correlated response can be either favorable or unfavorable. A favorable correlated response occurs when selection for one trait improves another trait that is also desirable. An unfavorable correlated response occurs when selection improves the target trait but causes deterioration in another economically or biologically important trait. This creates potential antagonistic genetic relationships that must be managed through balanced breeding objectives.
- An important example is the relationship between production and fitness traits. High production may have an unfavorable genetic association with fertility, health, survival, or longevity in some populations and production systems. If selection is based only on production, correlated deterioration in fitness traits may occur. Incorporating fertility, health, survival, welfare, and functional traits into the breeding objective can help maintain a more balanced pattern of genetic improvement.
- Indirect selection is closely related to correlated response. Indirect selection occurs when one trait is selected because it provides useful information about another trait that is difficult, expensive, late-expressed, sex-limited, or difficult to measure directly. For example, an indicator trait may be easier to measure than the target trait. If the indicator and target trait have a sufficiently strong genetic correlation, selection using the indicator can produce genetic improvement in the target trait.
- The usefulness of indirect selection depends not only on genetic correlation but also on the heritability and accuracy of measurement for the indicator trait. A highly heritable indicator trait with a strong genetic correlation to a difficult target trait may provide an efficient route to genetic improvement. This principle is important in traits such as disease resistance, feed efficiency, fertility, carcass traits, and other characteristics that may be expensive or difficult to measure routinely.
- Genomic selection has increased the ability to predict correlated genetic responses. Genomic estimated breeding values (GEBVs) can use information from thousands of genetic markers to improve the prediction of additive genetic merit. In multi-trait genomic selection, information from genetically correlated traits can contribute to the prediction of breeding values, particularly when some traits have limited phenotypic records. A well-designed reference population containing phenotypic and genomic information is important for achieving reliable genomic predictions.
- Modern BLUP and multi-trait genetic evaluation systems can account for genetic and environmental relationships among traits. A simplified animal model is y = Xb + Za + e, where y represents observations, b represents fixed effects, a represents random additive genetic effects, and e represents residual effects. Multi-trait models extend this framework by estimating genetic variances and covariances among several traits, allowing breeding values and expected correlated responses to be evaluated jointly.
- Genotype–environment interaction can further complicate correlated response. Genetic relationships between traits may differ among environments, production systems, or management conditions. For example, selection for growth under high-input conditions may not produce the same correlated response in growth, fertility, or survival under heat stress or low-input systems. Therefore, breeding programs intended for diverse environments may need to consider environment-specific genetic parameters, adaptation traits, and genetic correlations across environments.
- Correlated response is also connected with genetic trends and long-term genetic improvement. Monitoring breeding values across generations allows breeders to determine whether selection is producing the desired changes in target and correlated traits. A favorable trend in one production trait accompanied by an unfavorable trend in fertility, health, or longevity can indicate that the breeding objective or selection index needs adjustment.
- The management of correlated response is therefore an essential part of sustainable animal breeding. Breeding programs should not simply maximize short-term response in one economically important trait. They should consider the complete set of traits affecting profitability, animal health, welfare, reproduction, longevity, adaptation, and environmental sustainability. Appropriate selection objectives, economic weights, selection criteria, genetic correlations, and selection indexes help breeders achieve balanced genetic improvement.
- Correlated response is also relevant to maintaining genetic diversity. Strong selection pressure concentrated on a limited number of animals can increase the use of popular sires and accelerate genetic concentration, inbreeding, and loss of genetic variation. Breeding programs may therefore combine genetic improvement with management of inbreeding, optimal contribution selection, and mate allocation. Maintaining sufficient additive genetic variation is important because future genetic progress depends on the availability of useful genetic variation.
- The distinction between direct response and correlated response is fundamental. Direct response refers to genetic change in the trait under selection, whereas correlated response refers to genetic change in another trait resulting from the genetic relationship between the two traits. Understanding both responses allows breeders to predict the broader consequences of selection rather than evaluating improvement only for the trait used to rank animals.
- Overall, Correlated Response to Selection describes how selection on one trait can produce genetic change in another trait because of shared additive genetic effects and genetic covariance. Its magnitude and direction depend on genetic correlation, additive genetic variance, heritability, selection intensity, selection accuracy, and the information used in genetic evaluation. In modern animal breeding, understanding correlated response is essential for designing multiple-trait selection, selection indexes, genomic selection, and balanced breeding objectives that achieve genetic improvement while protecting fertility, health, welfare, longevity, adaptation, and genetic diversity.
