A fair lottery mechanism underlies self-determined caste development in Melipona bees

In Melipona stingless bees, larvae develop in sealed, identically sized, mass-provisioned cells. This enables larvae to determine their own caste fate, causing 5–15% of females to develop as queens, a frequency approaching the individual optimum but exceeding colony requirements. Self-determination has been suggested to operate through a stochastic switch, whereby intrinsic noise or environmental cues trigger alternative caste trajectories. This “fair lottery” hypothesis predicts a spatially random distribution of queens, whereas localised maternal or worker control should induce clustering. We tested this prediction using a 25-year dataset comprising 593 brood-comb maps from nine Melipona species. Multiscale spatial analyses showed that gyne positions were indistinguishable from random, while collectively controlled male production was clustered. Gyne production varied primarily among colonies and months and increased with comb cell number – a proxy for cell-building rate. These results suggest that Melipona caste determination resembles a fair lottery but with dynamically adjustable odds: within a contemporaneous brood cohort, larvae have an equal chance of developing as queens, but the colony-level odds change over time. Such recalibration is consistent with evolutionarily stable strategy theory, where seasonal variation in worker mortality and cell-building rates alters the colony-level cost, and hence the stable level, of queen overproduction.