Winter starts in August

The bees that carry a colony through winter are not the same animal as the ones flying now. A summer worker lives about six weeks and works herself to death. A winter bee lives for months, with enlarged fat bodies and high levels of a storage protein called vitellogenin, and her job is mostly to sit in a cluster shivering and stay alive until spring.

Those bees are reared in late summer and autumn, and that single fact drives the whole calendar. A colony carrying a heavy mite load while it rears winter bees produces winter bees that do not live long enough. The colony looks fine in September because the summer bees are still flying. It dies in January or February, when the bees that should have replaced them are already gone.

This is the answer to why a strong-looking hive dies over winter with capped honey still in the box. It did not starve and it did not freeze. It ran out of bees.

1. Mites first, and confirm it worked

Nothing else on this list rescues a colony that goes into autumn with high mites, so this is the one to do first and the one to be least casual about. Count with an alcohol wash or sugar roll, treat if you are over threshold, and pick a product that suits the temperatures you actually have.

Then count again two to three weeks later. That follow-up count is the step almost everyone skips, and it is the only way to know the treatment did anything. Miticides fail for ordinary reasons, including resistance, a product applied at the wrong temperature, and a dose sized for the wrong number of boxes. A treatment you did not verify is a treatment you are guessing about.

2. Stores, and where they sit

How much a colony needs varies enormously by climate, and the target for a long northern winter can be double what a mild one needs. That question has its own guide and its own calculator, so I will not repeat the numbers here.

What gets less attention is where the stores are. A winter cluster moves upward as it eats, and moves sideways only reluctantly and only when it is warm enough to break formation. A colony can starve with full frames of honey two inches to the left of the cluster. When you are setting up for winter, the question is not only "is there enough" but "is it above them".

The practical test is hefting. Lift the back of the hive an inch off the stand. A colony ready for winter feels solidly anchored to the ground. Do it every few weeks through autumn and again on mild days in winter, and the change over time will tell you far more than any one reading.

3. Feed while they can still take it

Autumn feed is 2:1 syrup, twice as much sugar as water by weight, because the bees have to evaporate less of it before they can cap it. Feed it early enough that they can finish the job. As temperatures drop, colonies stop taking syrup down and stop being able to cure it, and at that point you have moved from feeding to emergency feeding, which means fondant or sugar laid directly above the cluster.

If you are close to the line, err on the side of feeding sooner rather than more. Syrup they cannot process is not stores, it is a bucket of sugar water sitting in a cold hive.

4. Check the queen while there is still time to act

Autumn is the last point at which a queen problem is fixable. A colony that is queenless, or laying drone brood in a worker pattern, will not survive winter and will not fix itself. You want to know that in September, when combining is still an option, rather than in November, when it is not.

Look for eggs and a decent pattern of worker brood. The colony will be winding down laying, so a smaller brood nest than midsummer is normal and expected. What you are looking for is evidence that a laying queen exists, not a big nest.

5. Combine, rather than gamble on two weak colonies

Two weak colonies going into winter usually produce two dead colonies by spring. Combined, they often produce one that lives, and one live colony in April is worth more than two dead ones. Lay a sheet of newspaper between the boxes with a couple of small slits in it, put the weaker colony on top, and let them chew through slowly enough to merge without fighting.

This is the least sentimental decision in beekeeping and one of the most reliable. If you are hesitating over whether a colony is strong enough, that hesitation is usually the answer.

6. Moisture kills more colonies than cold

A cluster generates a surprising amount of water vapour. That vapour rises, meets the coldest surface in the hive, which is usually the underside of the lid, condenses there and drips back down onto the bees. Wet bees in a cold hive die. Dry bees survive temperatures that sound impossible.

This is where the wrapping argument starts, though it does not finish here. Insulation above the cluster does work nobody disputes: it stops the lid being the coldest surface, which is where water forms. Wall insulation is the contested part. The traditional view is that it matters much less. Derek Mitchell's thermal work at the University of Leeds cuts the other way, arguing that a thin wooden box is a poor thermal home compared with the tree cavities bees choose for themselves, and that the winter cluster is a stress response rather than the neat insulating layer it is usually described as.

Both can be acted on: insulate above the cluster, because that is the part everyone agrees about and it is cheap. Treat wall insulation as an open question worth reading into rather than a settled yes or no.

The other half is giving the moisture somewhere to go, whether that is a small upper entrance, a vented lid or an absorbent layer such as a quilt box. Which of those you use depends on your climate and your equipment, and beekeepers hold strong views. The principle underneath all of them is the same: do not let water collect above the cluster.

7. Mice, wind and the entrance

Mice move into hives when the nights turn cold and the bees stop patrolling the entrance, which is earlier than most people expect. Fit a mouse guard or reduce the entrance to something a mouse cannot pass before the first properly cold nights, not after you have seen evidence.

A reduced entrance also cuts the draught through the hive and gives a smaller colony less to defend against robbing, which is at its worst in autumn dearth. If your site is exposed, a windbreak on the prevailing side is worth more than a wrap.

8. Take the space away

Pull the supers you are not leaving for stores, and take out the queen excluder. A colony only has to keep its cluster warm, not the whole box, but empty space above the bees is space where moisture collects and heat is lost.

The excluder is worth singling out. The cluster moves up through winter; an excluder can hold the queen below while everyone else moves above her. It is an easy thing to forget in October and a miserable thing to find in March.

9. Know when to stop looking

Full inspections end when daytime highs settle below roughly 55°F (13°C). Below that you are breaking the cluster and chilling brood in exchange for information you could have had another way. From then on, the tools are hefting for weight, watching the entrance on mild days, and on a still cold day putting an ear against the side of the box.

If a colony is light in midwinter you can lift the lid briefly on a mild day to lay fondant directly over the cluster. That is a rescue, not an inspection, and it should take under a minute.

A rough timetable

Shift this by a few weeks in either direction for your latitude, and use your own first frost date as the anchor rather than the month names.

  • Late summer. Mite count, treat, re-count to confirm. Assess the queen. This is the month that decides the outcome.
  • Early autumn. Combine anything too weak to winter. Start 2:1 syrup if the colony is light. Remove supers you are not leaving.
  • Mid autumn. Finish feeding while they can still cure it. Mouse guard on, entrance reduced, excluder out, insulation above the cluster.
  • Winter. Heft every few weeks. Emergency feed only if light. Otherwise leave them alone.
  • Late winter. The dangerous stretch. Brood rearing restarts, consumption climbs sharply, and colonies that starve usually starve now rather than in December.

What actually goes wrong

Across every deadout I have looked at, the causes cluster into a short list, and it is not the list beginners expect. Mite-damaged winter bees come first by a distance. Starvation in late winter comes second, often with honey still in the box but out of reach. Damp comes third. Queen failure going into autumn comes fourth. Cold on its own is a long way down and is usually a symptom of one of the others.

Which is the useful thing about this checklist. The four items at the top are not four preferences. They are, in order, the four things that kill colonies.

Sources and further reading

Every month named above shifts with latitude. Anchor the timetable to your own first frost date and to what your local association is doing, not to the calendar.