A farm road in the far northwest corner of Washington, a few miles short of the Canadian line.
Before dawn on a Friday in late spring, a semitrailer takes a tight corner badly and goes over.
Its load is 70,000 pounds of beehives, and about 300 of them break open in the ditch.
The first number that reaches the public says 250 million bees are loose.
That figure lasted about a weekend.
The real one was 14 million.
Why a spilled colony walks back into a broken box
A loose swarm on a roadside is not a swarm in the ordinary sense. It is a colony that has lost its house and is trying to find it again.
What holds a colony together is one insect. The queen produces a chemical signature that every worker in the hive knows, and their whole orientation runs off it.
So the recovery method is not sweeping bees into anything. It is repairing the boxes, setting them upright near where they fell, and then waiting.
Workers in the air read the scent, find the frames their queen is on, and return on their own. Bees also mark the entrance with a scent of their own, which pulls in the ones still circling.
Move the boxes and you break the loop. Leave them and most of the colony comes home by the second morning.
The hive is the address. The queen is the reason.
Where 250 million came from
Nobody counted 250 million bees. Nobody could. The figure was an estimate made quickly at a chaotic scene and repeated by everyone downstream of it.
Here is the arithmetic that undid it. A loaded hive box in pollination service weighs somewhere around 80 to 100 pounds.
Divide 70,000 pounds by that and the trailer was carrying a few hundred hives, not thousands.
A strong colony in spring holds a few tens of thousands of workers. Multiply a few hundred hives by a few tens of thousands and 14 million falls out of it immediately.
To reach 250 million you would need something like a third of a million bees in every single box, which is roughly ten times what a hive can physically hold.
The load was weighed. The bees were guessed.
Who fixed it and how fast
The correction did not come from an agency review. It came from people in the area who did the division themselves and said so out loud.
More than two dozen beekeepers turned up alongside deputies and emergency crews, which in a county this agricultural is a normal Friday call out.
They rebuilt what boxes could be rebuilt and left the rest standing where the bees could reach them.
Most of the colonies were back inside by Monday morning. Several responders were stung and the driver walked away unhurt.
The sheriff’s office later acknowledged the people who had corrected its own arithmetic, which is more than most agencies do.
Two dozen beekeepers, three days, and one correction.
Why the wrong number travels further than the right one
A quarter of a billion insects is an irresistible headline, and 14 million is not, so the first figure had a day of open road before anything caught it.
That asymmetry is structural rather than malicious. The estimate arrives in the first hour, the correction arrives after somebody has done the work.
Nothing about the incident changes with the smaller number either, which is the part worth sitting with. It was still a large loss for the owner and still a serious recovery job.
The event mattered enough that the state ended up writing a response plan for exactly this kind of spill, a document that had not existed anywhere before.
That plan and the same crash are covered under a 10 page state procedure.
The corrected count and the recovery are reported by a museum magazine.
The first number was eighteen times the real one.
What actually decides whether the bees survive
Getting insects back into boxes is the visible half. Whether those colonies live is decided over the following weeks and nobody films that.
A colony that loses its queen in a crash will not recover on its own. It needs a replacement queen introduced, and introductions fail often.
Crushed comb is the other cost. Brood killed in the wreck is a generation that will not emerge, and the gap shows up a month later as a colony too small to work.
Commercial hives are also on a schedule. A load like this is usually moving toward a pollination contract, and a colony that recovers slowly misses the bloom it was rented for.
A similar rollover in Texas left a beekeeper expecting roughly a quarter of 408 hives to survive.
The cause, the hive damage and the responder injuries are described by a local station.
The honey bees came back to their boxes, and whether they made it is a different question with a slower answer.
Read the whole thing?
Get the week's signal, not the noise
Our sharpest reporting on energy, climate and nature — free, once a week.
