- Genetic covariance is the statistical association between genetic effects for two traits within a population. It describes whether genetic differences that influence one trait tend to be associated with genetic differences in another trait. Genetic covariance is a fundamental concept in quantitative genetics, evolutionary genetics, plant and animal breeding, and the study of complex traits.
- For example, if genetic factors that increase body weight also tend to increase growth rate, the two traits may have positive genetic covariance. If genetic factors that increase one trait tend to decrease another, the traits may have negative genetic covariance. If genetic effects influencing the two traits are largely independent, their genetic covariance may be close to zero.
- Genetic covariance is distinct from ordinary phenotypic covariance. Phenotypic covariance describes how two observed traits vary together, whereas genetic covariance describes how their genetic components vary together. Environmental factors can create phenotypic covariance even when the underlying genetic covariance is small.
- For two traits, genetic covariance can be represented as: Cov(A₁, A₂)
- where A₁ and A₂ represent the additive genetic values for traits 1 and 2. More generally, genetic covariance can be defined for different types of genetic effects, including additive, dominance, and epistatic effects.
- A positive genetic covariance means that individuals with higher genetic values for one trait tend to have higher genetic values for the other. A negative genetic covariance means that higher genetic values for one trait tend to be associated with lower genetic values for the other. A covariance near zero indicates little linear association between the genetic effects of the two traits.
- Genetic covariance is closely related to genetic correlation. Genetic correlation standardizes genetic covariance by the genetic standard deviations of the two traits: rᴳ = Cov(A₁,A₂) / √(Vᴬ₁Vᴬ₂)
- where rᴳ is the genetic correlation, Cov(A₁,A₂) is genetic covariance, and Vᴬ₁ and Vᴬ₂ are the additive genetic variances of the two traits. Genetic correlation ranges from −1 to +1.
- Covariance and correlation therefore provide related but different information. Genetic covariance depends on the scale and units of measurement of the traits, whereas genetic correlation is standardized and describes the strength and direction of their genetic association.
- Genetic covariance can arise when the same genes or genetic variants influence multiple traits. This is often associated with pleiotropy, in which a single genetic locus affects more than one trait. Genetic covariance can also arise because different genetic loci influencing two traits are correlated through linkage disequilibrium or population structure.
- For example, a genetic variant may influence both growth and metabolism. If individuals carrying genetic variants associated with increased growth also tend to have genetic variants associated with altered metabolic traits, the traits may show genetic covariance.
- Genetic covariance is an important component of genetic architecture. Complex traits are often influenced by many genes, and some of these genes may affect multiple phenotypes. Understanding covariance among traits therefore helps researchers determine how different characteristics are genetically connected.
- Genetic covariance can be positive or negative. Positive covariance can cause two traits to respond in the same direction to selection, while negative covariance can create a genetic trade-off. These trade-offs are particularly important when improving one trait has unfavorable consequences for another.
- In animal breeding, genetic covariance is important when breeders select for multiple economically important traits. For example, selection for increased production may be genetically associated with changes in fertility, disease resistance, growth, or longevity. Understanding the genetic covariance between these traits helps breeders predict correlated responses.
- The relationship between genetic covariance and selection can be described using the multivariate breeder’s equation:
- Δz = Gβ
- where Δz represents the vector of expected changes in traits, G is the G-matrix, and β represents selection gradients. The G-matrix contains additive genetic variances on its diagonal and additive genetic covariances between traits off the diagonal.
- The G-matrix, also called the additive genetic variance–covariance matrix, is one of the central tools of multivariate quantitative genetics. It summarizes the amount of additive genetic variation within each trait and the genetic covariance among traits.
- For example, a simplified G-matrix for two traits can be written as:
- G = [ Vᴬ₁ Cov(A₁,A₂) ]
- ** [ Cov(A₁,A₂) Vᴬ₂ ]**
- The diagonal elements represent additive genetic variances, while the off-diagonal elements represent additive genetic covariance. For multiple traits, the matrix expands to include the genetic variances and covariances of all traits.
- The G-matrix helps researchers predict how a population is expected to respond to natural selection or artificial selection. If two traits have strong positive genetic covariance, selection favoring one trait may produce an increase in the other. If they have strong negative covariance, selection for one may cause an unfavorable change in the other.
- Genetic covariance therefore provides a foundation for understanding correlated response to selection. A breeder may select directly for one trait but observe changes in another trait because the two traits share genetic influences.
- Genetic covariance can also influence evolutionary trajectories. Natural selection rarely acts on only one trait in isolation. Organisms have multiple characteristics that interact with survival and reproduction. Genetic covariance among these traits can constrain or facilitate evolutionary change.
- A population may have strong selection favoring an increase in one trait, but if that trait has a strong negative genetic covariance with another trait that is also important for fitness, evolutionary change may be constrained. Conversely, positive genetic covariance can allow coordinated evolutionary change.
- Genetic covariance is therefore closely connected to evolutionary constraints and adaptive evolution. The structure of genetic covariance can influence which combinations of traits are readily produced by genetic variation and which evolutionary directions are more difficult.
- Genetic covariance can be estimated from covariance among relatives. Family members share genetic material, so the degree to which relatives resemble one another for multiple traits can provide information about genetic relationships between those traits. Parent–offspring, full-sib, half-sib, and pedigree-based designs have historically been used for this purpose.
- Pedigree analysis can estimate genetic relationships among individuals and help separate genetic covariance from environmental covariance. More recently, genomic relationship matrices have allowed researchers to estimate genetic relationships using genome-wide genetic markers.
- Genomic methods can be particularly useful when large populations have been genotyped. Genome-wide markers can help estimate realized genetic relationships and improve estimates of genetic covariance among traits.
- Genetic covariance can also be investigated using QTL mapping and genome-wide association studies (GWAS). If genomic regions affect multiple traits, they may contribute to genetic covariance through pleiotropic effects or linked genetic variants.
- However, observing that the same genomic region is associated with multiple traits does not necessarily prove biological pleiotropy. The associations may arise from linkage disequilibrium, correlated environmental effects, population structure, or statistical relationships between traits. Careful experimental and statistical analysis is required.
- Genetic covariance is influenced by allele frequencies and population structure. A genetic variant can contribute differently to covariance in different populations because its frequency and association with other variants may differ. Consequently, genetic covariance estimates are population-specific.
- Environmental conditions can also influence the observed relationship between traits. Genotype–environment interaction (G×E) occurs when genetic effects differ across environments. As a result, genetic covariance between traits can sometimes change depending on environmental conditions.
- For example, two traits may have strong positive genetic covariance under one environmental condition but weaker covariance under another. This can occur when environmental factors alter gene expression, physiological pathways, or the relative importance of different genetic variants.
- Genetic covariance is also related to pleiotropy. Pleiotropic genes can influence multiple traits because they participate in shared biological pathways or have effects in different tissues or developmental stages. Pleiotropy can therefore create correlations among traits at the genetic level.
- Another mechanism is linkage disequilibrium, where alleles at different loci occur together more or less frequently than expected by chance. If different loci influence different traits and their alleles are correlated in the population, linkage disequilibrium can contribute to observed genetic covariance.
- Epistasis may also contribute to genetic covariance when interactions between loci influence multiple traits. Such interactions can make the genetic relationships among traits more complex than predicted by simple additive models.
- Genetic covariance is particularly important for traits that are biologically connected. Growth, body size, development, metabolism, reproduction, and disease resistance may share underlying genetic pathways. The resulting covariance can influence both breeding outcomes and evolutionary responses.
- In plant breeding, genetic covariance can be used to understand relationships among yield, flowering time, plant height, drought tolerance, disease resistance, and other agronomic characteristics. A breeder can use the genetic covariance structure to design selection strategies that improve several traits simultaneously.
- In animal breeding, genetic covariance is important for evaluating traits such as growth, milk production, fertility, feed efficiency, disease resistance, and longevity. Selection programs often need to balance multiple traits rather than maximize a single characteristic.
- Genetic covariance also plays a role in genomic selection. Genomic prediction models can incorporate genetic relationships among traits and use information from genetically correlated traits to improve prediction accuracy. This is particularly useful when one trait is difficult or expensive to measure.
- In human genetics, genetic covariance is used to investigate relationships between complex traits and diseases. Genome-wide data can be used to estimate the extent to which genetic factors contributing to one trait overlap with those contributing to another.
- Such analyses can help researchers investigate shared genetic architecture among traits, although statistical genetic covariance should not automatically be interpreted as proof that one trait causes another. Genetic association and biological causation are different concepts.
- Genetic covariance should also be distinguished from phenotypic covariance and environmental covariance. Two traits can covary phenotypically because they respond to the same environmental conditions even if their genetic components are independent. Conversely, genetic covariance can exist even when environmental effects obscure the relationship at the phenotypic level.
- For this reason, quantitative genetic models attempt to partition observed covariance into genetic and environmental components. Linear mixed models, multivariate animal models, pedigree models, and genomic models can be used to estimate these components.
- The precision of genetic covariance estimates depends on sample size, relationship structure, trait measurement quality, population structure, environmental variation, and the statistical model. Estimates based on small datasets may have substantial uncertainty.
- Genetic covariance is also important when considering heritability. Each trait has its own additive genetic variance and heritability, but the covariance between traits determines how selection on one trait may influence another. Therefore, heritability alone is not sufficient to predict responses when selection involves multiple traits.
- The combination of genetic variances and covariances provides a more complete picture of the genetic structure of a population. This is why the G-matrix is central to multivariate quantitative genetics and evolutionary biology.
- Overall, genetic covariance describes how genetic effects influencing different traits vary together. It connects genetic variance, heritability, genetic correlation, pleiotropy, linkage disequilibrium, and the G-matrix with the practical problems of selection and the evolutionary problem of correlated trait change.
- Understanding genetic covariance is essential for predicting how populations respond to selection, identifying genetic relationships among complex traits, designing breeding programs, and studying evolutionary constraints and adaptation. Together with genetic variance and covariance among relatives, it provides a powerful framework for understanding the genetic architecture of multiple traits.
