Building a Selective Breeding Programme: Heritability and Selection Indices for Honeybee Traits
How UK breeding groups use heritability estimates and simple selection indices to make measurable genetic progress on traits like hygienic behaviour, gentleness and productivity.
Why individual selection alone is slow
Most beekeepers who want to improve their stock do it informally: they notice a colony that produces well, is calm to work, and overwinters strongly, and they raise queens from it the following year. This works, but progress is slow and inconsistent because a single colony's performance in a single season is a noisy signal, affected by weather, forage, disease pressure and simple luck as much as by genetics. Formal breeding programmes, run by groups such as regional bee improvement and bee breeders' groups in the UK, instead track performance data across many colonies and several years before deciding which lines to propagate, which produces far more reliable genetic gain.
The bees' unusual reproductive biology adds a complication that plant and livestock breeders do not face: a queen's genetic contribution to her workers is only half the story, because she typically mates with 10-20 drones from the surrounding area, so a colony's worker population is genetically a patchwork of many half-sister subfamilies. This means colony-level performance reflects both the queen's genes and an uncontrolled sample of the local drone population, unless mating is deliberately controlled through instrumental insemination or isolated mating stations.
Heritability: how much of what you see is genetic
Heritability is a statistic, expressed as a value between 0 and 1, describing what proportion of the variation seen in a trait across a population is due to genetic differences rather than environment or chance. It is not a fixed property of a trait in the abstract, since it depends on the population and conditions being measured, but published estimates give useful working guidance. Traits like defensive behaviour and hygienic behaviour tend to show moderate-to-high heritability, meaning selection can move the needle relatively quickly, while traits like honey yield show more moderate heritability because they are heavily influenced by forage and weather in any given year, diluting the genetic signal.
A practical consequence is that traits with lower heritability need to be measured over more colonies and more years before a breeder can be confident a particular queen line is genuinely superior rather than simply lucky. This is why serious breeding groups insist on multi-year performance recording rather than promoting a queen line to breeding stock after one exceptional season, and why single-season 'best hive' selection, however tempting, produces much less reliable long-term improvement than sustained record-keeping.
Building a simple selection index
A selection index combines scores on several traits into one overall ranking, weighted by how important each trait is to the breeder's goals and how reliably it can be measured. A hobbyist-friendly version might score each colony from 1 to 5 on temperament (calmness on the comb, following behaviour), productivity (honey stored relative to colony size), hygienic behaviour (speed of removing freeze-killed brood in a standard test), and overwintering (colony strength in spring relative to strength the previous autumn), then combine these into a weighted total that reflects the breeder's priorities, for example weighting temperament and hygienic behaviour more heavily than raw honey yield if disease resilience and safety around neighbours matter more locally.
The freeze-kill test for hygienic behaviour is a good entry point for hobbyist breeders because it is cheap, quick and reasonably reliable: a patch of sealed brood is frozen in situ using a section of comb dipped in liquid nitrogen or a similar method, then the colony is scored on the percentage of dead brood it has uncapped and removed within 24-48 hours. Colonies that clear more than roughly 95% of a standard test patch within two days are generally considered strongly hygienic, and hygienic behaviour correlates well with resistance to both chalkbrood and, to a useful degree, varroa, since bees detect and remove mite-infested pupae through the same uncapping-and-removal mechanism.
Managing inbreeding and genetic diversity
A breeding programme that selects too narrowly, propagating queens from just one or two outstanding colonies year after year, risks a genetic bottleneck that shows up as diploid drones, a spotty brood pattern caused by the queen's inability to produce viable diploid females at certain sex-allele combinations, rather than disease or a failing queen. Honeybees have a single-locus complementary sex determination system, meaning a fertilised egg only develops into a female worker if it is heterozygous at the sex locus; homozygous fertilised eggs develop into infertile diploid drones that are eaten by workers shortly after hatching, producing a shot-gun brood pattern that looks alarmingly like disease but is actually a genetic diversity problem.
Maintaining several distinct breeding lines, exchanging queens or drone semen with other breeding groups periodically, and avoiding repeatedly mating a line's queens with drones from closely related colonies are the standard safeguards. Regional bee improvement groups in the UK often coordinate isolated mating apiaries, sited well away from other beekeepers' colonies, specifically so that queens raised from selected breeder colonies mate with a known and deliberately diverse pool of drones rather than an uncontrolled local population.
Frequently Asked Questions
What is a good heritability estimate to know as a starting breeder?
As general working figures drawn from published apicultural research, temperament and hygienic behaviour tend to sit in a moderate-to-high heritability range, while honey yield tends to be more moderate, meaning yield needs more years of data before selection decisions can be made with confidence.
How do I know if my colony's spotty brood pattern is genetic rather than disease?
A shotgun brood pattern caused by diploid drone mortality typically shows scattered, isolated empty cells within an otherwise healthy-looking patch surrounded by capped brood, with no discolouration, odour or sunken, perforated cappings; those signs point instead to a brood disease and should be checked against a professional diagnosis rather than assumed to be genetic.
Do I need instrumental insemination to run a breeding programme?
No. Most amateur and regional group breeding programmes rely on open mating at isolated apiaries or by controlling the local drone population through drone-source colonies, reserving instrumental insemination for research institutions and a small number of specialist commercial breeders who need guaranteed paternity.
How many colonies do I need to run a meaningful selection programme?
Serious selection gains typically require comparing at least a dozen or more colonies over two or more seasons; with only two or three hives you can still record and compare performance usefully, but drawing firm genetic conclusions from such a small sample is unreliable.