- Heritability is a fundamental concept in quantitative genetics that describes the proportion of variation in a trait within a particular population and environment that is associated with genetic differences among individuals. It is especially important for understanding why individuals differ in complex traits and for predicting how traits may respond to selection. Heritability does not describe how “genetic” an individual trait is, nor does it measure the proportion of a person’s phenotype that is caused by genes. Instead, it is a population-level statistical property that depends on the genetic and environmental conditions in which a trait is measured.
- A phenotype can be represented as the combined result of genetic and environmental influences. In a simplified quantitative-genetic model, phenotypic variance is expressed as VP=VG+VEV_P = V_G + V_E, where VPV_P is phenotypic variance, VGV_G is genetic variance, and VEV_E is environmental variance. Genetic variance can be further divided into additive genetic variance, dominance variance, and epistatic variance, giving VG=VA+VD+VIV_G = V_A + V_D + V_I in a more detailed model. These variance components provide the foundation for understanding the different forms of heritability.
- Broad-sense heritability, denoted H2H^2, is the proportion of phenotypic variance attributable to total genetic variance. It is calculated as H2=VG/VPH^2 = V_G/V_P. Because VGV_G can include additive, dominance, and epistatic effects, broad-sense heritability captures the overall contribution of genetic differences to phenotypic variation under the conditions studied. It can therefore be useful for understanding the genetic contribution to complex traits, particularly when the objective is to describe total genetic variation rather than predict the response to selection.
- Narrow-sense heritability, denoted h2h^2, is the proportion of phenotypic variance attributable specifically to additive genetic variance. It is calculated as h2=VA/VPh^2 = V_A/V_P. Narrow-sense heritability is particularly important in evolutionary genetics and animal and plant breeding because additive genetic effects are transmitted from parents to offspring in a way that allows them to contribute predictably to the response to selection. For this reason, narrow-sense heritability is often the more relevant measure when predicting genetic response to selection.
- The distinction between broad-sense and narrow-sense heritability is closely related to the distinction between genetic variance and additive genetic variance. Total genetic variance can include genetic interactions that do not necessarily contribute predictably to offspring resemblance in the same way as additive effects. Dominance variance results from interactions between alleles at the same locus, while epistatic variance results from interactions between alleles at different loci. These components can contribute substantially to phenotypic differences without being equivalent to additive genetic variation.
- Heritability is closely connected to breeding value, which represents the additive genetic contribution of an individual that is expected to be transmitted to its offspring. Individuals with high breeding values for a desirable trait may be preferentially selected as parents in breeding programs. Because breeding values are based on additive genetic effects, narrow-sense heritability is central to predicting how strongly selection on phenotypes will produce genetic change across generations.
- The relationship between heritability and selection is commonly expressed through the breeder’s equation, R=h2SR = h^2S, where RR is the response to selection, h2h^2 is narrow-sense heritability, and SS is the selection differential. The equation shows that the expected response depends on both the amount of selection applied and the proportion of phenotypic differences associated with additive genetic differences. When narrow-sense heritability is higher, the same selection differential generally produces a larger expected response, assuming the underlying assumptions of the model are satisfied.
- Heritability is therefore not a fixed property of a trait or species. It can change among populations, environments, generations, and experimental conditions. For example, if environmental conditions become more uniform, environmental variance may decrease, causing heritability to increase even though the underlying genetic effects have not changed. Conversely, greater environmental heterogeneity can increase environmental variance and reduce heritability. A change in heritability does not necessarily mean that genes have become more or less important biologically.
- This point is important when interpreting heritability in human genetics. A highly heritable trait is not necessarily unaffected by environmental conditions, and a trait with low heritability is not necessarily unimportant genetically. Heritability describes variation among individuals within a specific population and environment. It does not indicate how much of an individual’s phenotype is caused by genes, and it cannot by itself determine whether a trait can be changed by modifying the environment.
- Environmental variance is therefore an essential component of heritability. Factors such as nutrition, temperature, disease exposure, management, education, socioeconomic conditions, developmental conditions, and measurement error can contribute to phenotypic variation. When environmental effects differ among individuals, they can reduce the proportion of total phenotypic variance explained by genetic differences. Understanding environmental variance is consequently essential for correctly interpreting estimates of heritability.
- Genotype–environment interaction (G×E) adds another layer of complexity. Different genotypes may respond differently to environmental conditions, meaning that the ranking or magnitude of genetic effects can change across environments. In such situations, heritability may differ between environments, and a single estimate may not adequately describe the genetic architecture of the trait across all conditions. G×E is particularly important in agricultural breeding, where varieties or animal genotypes may perform differently across locations, climates, diets, or management systems.
- Heritability can be estimated using several approaches. Covariance among relatives is a classical method in which the resemblance between parents and offspring, siblings, half-siblings, twins, or other relatives is used to infer components of genetic variance. Pedigree-based methods use an additive relationship matrix to model expected genetic relationships among individuals. Modern genomic approaches can instead use a genomic relationship matrix, constructed from DNA markers, to estimate genetic variance and heritability from genome-wide genetic similarity.
- Statistical models such as linear mixed models and restricted maximum likelihood (REML) are widely used to estimate genetic and environmental variance components. These approaches can incorporate fixed effects, random genetic effects, environmental factors, repeated measurements, pedigree relationships, and genomic information. The resulting estimates can be used to calculate broad-sense or narrow-sense heritability depending on the variance components included in the model.
- Heritability is closely related to genetic covariance and genetic correlation when multiple traits are studied simultaneously. Genetic covariance measures how genetic effects influencing two traits vary together, while genetic correlation standardizes this relationship by the genetic variances of the two traits. A trait may have moderate heritability while being strongly genetically correlated with another trait, allowing selection on one trait to produce a correlated response in another.
- The interpretation of heritability also depends on genetic architecture. Complex traits may be influenced by thousands of genetic variants, each with relatively small effects, as well as variants with larger effects. Polygenic inheritance, pleiotropy, dominance, epistasis, and linkage disequilibrium can all influence the distribution of genetic variation underlying a trait. Heritability summarizes the contribution of genetic variance at the population level but does not by itself reveal which genes or variants are responsible.
- Modern genetic studies can combine heritability estimates with QTL mapping, GWAS, sequencing, and genomic prediction to investigate the genetic basis of complex traits. Genome-wide association studies can identify genetic variants associated with trait differences, while heritability analysis asks a broader question about how much of the observed phenotypic variation is associated with genetic differences overall. The two approaches therefore provide complementary information.
- Heritability is also important in genomic selection, where genome-wide genetic markers are used to predict breeding values. Estimates of additive genetic variance and heritability help determine how much phenotypic information and genomic information may contribute to prediction accuracy. Genomic estimated breeding values (GEBVs) can then be used to rank individuals for selection, particularly in livestock, crop, forestry, and aquaculture breeding programs.
- In evolutionary genetics, heritability helps explain the potential for populations to respond to natural selection. If phenotypic differences are associated with additive genetic differences, selection can lead to changes in the distribution of traits across generations. However, evolutionary response depends on more than heritability alone. Genetic covariance among traits, mutation, migration, genetic drift, population structure, selection intensity, and environmental conditions can all influence evolutionary change. The multivariate response to selection is therefore better described using the G-matrix, which contains additive genetic variances and covariances among traits.
- Heritability is also relevant to genetic gain, artificial selection, and the design of breeding programs. Traits with substantial additive genetic variation and favorable selection conditions may respond efficiently to selection. However, breeders must also consider generation interval, selection intensity, population size, inbreeding, genetic correlations, and possible unfavorable correlated responses. High heritability alone does not guarantee rapid genetic improvement.
- A common misconception is that a heritability of 0.70 means that 70% of an individual’s trait is genetic. This interpretation is incorrect. A heritability of 0.70 means that, under the particular population and environmental conditions studied, approximately 70% of the observed phenotypic variance is associated with the relevant genetic variance represented by the chosen heritability measure. It does not mean that 70% of an individual’s phenotype is caused by genes.
- Another important limitation is that heritability estimates cannot automatically be transferred from one population or environment to another. Differences in allele frequencies, genetic diversity, environmental conditions, management, measurement methods, and population structure can all change the variance components and therefore the estimated heritability. Heritability should always be interpreted in the context of the population, environment, sample, and statistical model used to estimate it.
- Overall, heritability provides a quantitative framework for understanding the relationship between genetic and phenotypic variation. Broad-sense heritability describes the proportion of phenotypic variance associated with total genetic variance, whereas narrow-sense heritability focuses specifically on additive genetic variance and is particularly important for predicting response to selection. Together with genetic variance, environmental variance, covariance among relatives, breeding value, genetic covariance, and genetic correlation, heritability forms a central part of quantitative genetics and provides a foundation for studying inheritance, breeding, natural selection, and the genetic architecture of complex traits.
- Heritability is essential for predicting response to selection, estimating breeding values, understanding genetic architecture, designing breeding programs, interpreting family resemblance, and studying evolutionary change. It is most informative when considered together with the underlying genetic and environmental variance components rather than as a standalone measure of how genetic a trait is.
