Each hive in the grid represents one colony in a breeding population, scored on two traits: gentleness (calm, low-sting behaviour) and honey yield. Hive colour encodes gentleness (blue = defensive, green = gentle) and hive height encodes yield. Every generation, the breeder ranks colonies by a selection index — a weighted combination of both traits — and breeds only from the top scorers.
R = h² × S, where S is the selection differential (how far above average the selected parents are).Real honeybee breeding programmes (e.g. VSH and Carniolan gentleness lines) can take 5–10 generations to meaningfully shift a low-heritability trait like defensive behaviour, since colony-level traits mix many workers' genotypes and depend heavily on the queen's multiple matings.
A grid of 64 simulated honeybee colonies, each coloured by gentleness and sized by honey yield, evolves generation by generation as you tune heritability, selection intensity, and the selection index's trait weighting.
The breeder's equation R = h² × S in action: how strongly a trait responds to selection depends on both its heritability and how selective the breeder is, and how blending two traits in an index slows progress on each versus optimising alone.
Set heritability for gentleness and yield, choose a selection intensity, and slide the index weight toward yield or gentleness. Click "Breed next generation" repeatedly to watch the population's colour and height shift over time.
Because a colony's traits blend the genetics of a queen and many drones from multiple matings, honeybee breeding programmes often need far more generations than single-mated livestock to fix a desired trait.