Heritability Estimation for Production Traits

Heritability estimates what fraction of the observed variation in a trait across a population is attributable to genetic differences between individuals versus environmental and management factors, expressed as a value between zero, meaning essentially no observed variation is genetically driven, and one, meaning nearly all observed variation traces to genetic differences, a foundational parameter that directly determines how much genetic improvement a breeding program can realistically expect to achieve for a given trait through selective breeding alone. Traits with high heritability estimates, often including growth rate and some body conformation traits in several farmed species, respond relatively efficiently to straightforward selective breeding since a larger share of the observed performance differences between individuals genuinely reflects underlying genetic differences a breeding program can select for and pass on, while low-heritability traits, commonly including survival and various disease-resistance and fitness-related traits where environmental and management factors contribute disproportionately to observed variation, require either larger selection populations, more sophisticated statistical methods such as genomic selection that can extract genetic signal more efficiently than simple phenotypic selection, or dedicated controlled challenge-test phenotyping specifically to reduce confounding environmental variation and isolate the genuine underlying genetic signal more clearly. Heritability estimates are specific to the population, environment, and measurement conditions under which they were calculated, meaning a heritability value estimated in one breeding population or production environment should not be assumed to transfer precisely to a genetically different population or a meaningfully different production environment without independent re-estimation, since the same trait can show different heritability values under different environmental variance conditions even holding the underlying genetic architecture constant. Accurate heritability estimation requires adequately sized, well-structured family data with reliably recorded phenotypes and known pedigree or genomic relationships, meaning poor-quality underlying pedigree or phenotype data, the same data-integrity concern relevant to pedigree management software and genomic prediction generally, directly undermines heritability estimate reliability and, consequently, every subsequent breeding decision informed by an unreliable heritability figure.