The bees left, and then they came back, and they were different. This is the short version. The long version takes seventy years and raises more questions than it answers.

Colony collapse in the Great Lakes Metropolitan Zone reached terminal velocity around 2120. By 2125, wild honeybee populations were functionally extinct in the region. Managed colonies survived in controlled environments — sealed apiaries with filtered air and curated forage — but wild pollination ceased. The ecological consequences were severe and well-documented: crop failures in the agricultural zones, wildflower meadow die-offs in the wasteland margins, cascading declines in insectivorous bird populations. The bees were gone, and everything that depended on the bees began to fail.

Recovery began in 2186, when apiarists in the outer districts of GLMZ reported wild swarms establishing colonies in structural voids, abandoned HVAC systems, and the walls of derelict buildings. By 2200, wild colonies were common enough to map. By 2210, they were abundant. By 2224, the GLMZ Urban Apiary Collective estimates a wild honeybee population in the metropolitan zone of approximately 50 million individuals distributed across 2,000–3,000 colonies. The recovery is, by any measure, remarkable. The bees that recovered are, by any measure, not the bees that collapsed.

The new hive architecture is the most visually striking difference. Historical honeybee comb follows a regular hexagonal pattern — the most efficient tessellation for maximizing storage volume relative to wax expenditure. The new bees build comb that incorporates hexagonal cells but also pentagonal, heptagonal, and irregularly shaped cells in patterns that are geometrically complex and, according to structural analysis by the GLMZ Engineering Division, mechanically superior to standard hexagonal comb. The comb is stronger. It distributes load more efficiently. It contains internal channels that appear to serve ventilation and thermal regulation functions that standard comb does not. The bees have, through whatever process drives their construction behavior, invented a better architecture. How they arrived at it is unknown. Bees do not iterate. Bees do not prototype. Bees build what their genetics tell them to build. These bees are building something their genetics should not know.

The waggle dance has new movements. Honeybee communication through dance is one of the most studied phenomena in behavioral biology — the vocabulary is well-mapped, the grammar well-understood. The new bees perform the standard dance repertoire and also produce movement patterns that apiarists cannot decode. The new dances are consistent — the same patterns appear across multiple colonies — which argues against random variation. They are communicating something. We do not know what.

And the honey is wrong. Not wrong in the sense of harmful — toxicological testing consistently clears it for consumption, and it is, by all accounts, excellent honey. But spectroscopic analysis reveals trace compounds — complex organic molecules — that do not correspond to any plant species growing within the bees' documented foraging range. The bees are visiting flowers. Some of those flowers do not appear to exist in any botanical survey of the region. Where are the bees going? The Apiary Collective has attempted to track foraging scouts using miniaturized GPS transponders. The transponders lose signal within 3 kilometers of the hive, in a direction that leads toward the wasteland. The scouts return. The transponders do not resume transmitting until the scouts are back within range. Whatever is out there, wherever the bees are going, is in a dead zone. They go where we cannot follow, and they come back with honey that shouldn't exist.