Heritability in Honey Bees: What Selective Breeding Can and Cannot Fix

What heritability actually measures

Heritability is a statistic, not a fixed property of a trait carved in stone. It describes the proportion of the variation seen in a population, not in a single bee, that can be attributed to genetic differences rather than to environment, management, or chance. A heritability figure of 0.4 for a trait means that, within the specific population and conditions it was measured in, 40% of the observed variation between colonies tracks with genetic differences, and the remaining 60% comes from forage availability, weather, hive management, disease pressure, and pure noise.

This distinction matters enormously for breeders. A trait can be highly heritable and still be poor at predicting an individual colony's future performance, because heritability describes population-level variance, not a guarantee about any one colony. It also means heritability estimates are population- and environment-specific: a figure measured in a Mediterranean climate with abundant year-round forage does not automatically transfer to a heather moor apiary in the Scottish Highlands.

Traits with relatively high heritability

Honey bee colour and cuticle pigmentation patterns are among the most heritable traits available to breeders, which is why colour has historically been used, somewhat crudely, as a marker for subspecies and hybrid identification even though it tells you little about temperament or productivity. Tongue length, a trait of real historical interest for breeders targeting red clover pollination, also shows moderately high heritability and was one of the traits most successfully shifted by twentieth-century selection programmes.

Defensive behaviour (stinging response) and hygienic behaviour, the tendency of workers to detect and remove diseased or Varroa-infested brood, both show moderate to fairly high heritability in controlled breeding trials, which is precisely why hygienic behaviour has become the flagship trait in Varroa-tolerance breeding programmes worldwide, including UK-based queen rearing groups working with VSH (Varroa Sensitive Hygiene) lines. Because these behavioural traits are polygenic (controlled by many genes of small effect rather than one dominant gene), progress through selection is real but gradual, typically requiring several generations of testing and re-selection to shift a population meaningfully.

Traits with low heritability

Honey yield is the trait every beekeeper cares about most and, frustratingly, one of the least heritable in the practical sense, because it is driven overwhelmingly by forage availability, weather during the flow, colony size at the start of the season, and disease status, factors largely outside genetic control. A colony from excellent genetic stock in a poor forage year will often be outproduced by an average colony in a good year and location. This is why serious breeding programmes measure yield only as a relative comparison between colonies kept side by side in the same apiary under identical conditions, never as an absolute figure compared across different sites or years.

Overwintering survival is similarly muddied by heavy environmental influence, Varroa load, disease exposure, stores left by the beekeeper, and winter weather all swamp any underlying genetic contribution to hardiness, though there is growing UK interest in locally-adapted 'native' black bee (Apis mellifera mellifera) populations that appear to overwinter more efficiently on smaller stores, a trait that may have a genuine heritable component tied to broader physiological adaptation to a cooler, wetter climate.

How UK breeders apply this in practice

Practical selective breeding in a small UK operation rarely resembles the controlled, replicated trials used to generate published heritability estimates, but the underlying logic still applies. Breeders typically select breeder queens from colonies that have shown consistently good traits, calm temperament on the comb, low tendency to follow after a smoker puff, good hygienic behaviour tested with a simple pin-kill or liquid-nitrogen freeze-kill brood assay, and reasonable productivity relative to neighbouring colonies in the same apiary, over at least two seasons rather than one, since a single good season can be luck as much as genetics.

Instrumental insemination and controlled mating apiaries (isolated by distance or timing) give the most reliable genetic progress because they remove the randomness of open mating with unknown drones, but for hobbyist breeders working with open mating, the practical approach is to flood a local area with drones from your best colonies over several seasons, cull poorly performing colonies rather than breed from them, and accept that progress will be slower and noisier than in a controlled programme.

Environmental factors that mask genetic potential

Even a genetically excellent colony can look mediocre if it is managed poorly, so any breeding programme needs a way to separate signal from noise. Queen age is a major confound, since a colony's apparent temperament and productivity often shift as a queen ages and her pheromone output and laying rate decline, independent of her genetics. Local forage and weather in a given year can mask or exaggerate genetic differences between colonies to the point that single-season assessments are close to useless for breeding decisions. Disease and pest pressure, particularly uneven Varroa exposure between colonies in the same apiary, can make a genetically hygienic colony look average if it happens to sit next to a heavily infested colony that keeps reinfesting it through drift and robbing.