Every beekeeper eventually opens a hive on a mild late-winter day and finds silence. The temptation is to pull the frames, shake out the dead bees, and get the boxes cleaned up before the wax moths find them. Resist that for an hour. A dead colony is a crime scene, and the position of the cluster, the state of the brood, and what is lying on the bottom board will tell you more about your management than any live inspection ever did. The same cause tends to repeat if you do not name it, and most dead-outs in a small apiary trace back to three or four very recognizable patterns.

This guide walks through a dead-out the way an inspector would: outside first, then the cluster, then the stores, then the brood, then the debris. Keep a notebook or your phone open as you go. The findings only pay off if they change what you do next August.

Before You Touch Anything

Do the autopsy on the first day you discover the loss, before scavengers and robbing bees rearrange the evidence. Pick a dry day above about 40°F so you can work slowly without worrying about chilling anything in neighboring hives. Bring a hive tool, a bright flashlight, a white tray or sheet of cardboard, and a pair of reading glasses or a hand lens. You will be looking at things the size of a sesame seed.

Work from the outside in and write down what you see at each stage before forming a theory. Beekeepers who decide "it was varroa" at the entrance tend to stop looking, and they miss the isolation starvation or the drone-laying queen that was the actual killer. Log the date of the last live inspection, the last mite count, the last treatment, and the weight or stores estimate from fall. If you kept per-hive records in HiveBook or a paper log, pull them up now, because a dead-out is only interpretable against what the colony looked like going into winter.

A dead-out with no fall mite count is a guess. A dead-out with an August wash of nine mites per hundred bees is a diagnosis.

Step One: The Entrance and the Outside

Start at the landing board and the ground in front of the hive. A pile of dead bees outside the entrance in winter is normal to a point; undertaker bees drag out the dead on warm days. A large drift of dead bees with nothing left inside suggests the colony died slowly over weeks, with workers able to carry out corpses until the last few hundred could not. A hive with almost no dead bees outside but a solid carpet inside died fast, usually in a single cold snap.

Look at the front of the hive for brown or yellow streaking. Heavy fecal spotting on the entrance and front face points to dysentery, which in turn points to either Nosema or a long cold confinement on poor-quality stores such as honeydew or fermented syrup. Check the entrance reducer and bottom board for chewed comb and shredded debris, which mean mice got in. A mouse nest in the bottom box is rarely the cause of death on its own, but mice stress a cluster, foul the comb, and often arrive after the colony was already too weak to defend itself.

Note the hive weight by hefting the back. A hive that still feels heavy died with food in it, which rules out general starvation immediately and shifts suspicion to varroa, queen loss, or isolation starvation. A hive you can lift one-handed almost certainly starved, and the remaining question is whether that was your fault for underfeeding or the bees' fault for being too small a cluster to reach food.

Step Two: Find the Cluster and Measure It

Crack the inner cover and look down before pulling any frames. Where is the cluster, and how big is it? A healthy winter cluster in a northern climate starts around the size of a basketball in November and should still be larger than a grapefruit in February. A dead cluster the size of a softball or smaller, scattered across two or three frames, is the single most common dead-out picture, and it almost always means the colony went into winter with too few long-lived bees. The usual reason is varroa damage to the fall brood that should have become winter bees.

Now look at how the bees died. Pull the center frame of the cluster and hold it to the light:

  • Bees head-first in cells, abdomens sticking out — classic starvation. The bees crawled into empty cells reaching for the last traces of honey. If there is capped honey two frames over, this is isolation starvation: the cluster could not break contact to move to food during a cold stretch.
  • Bees in a tight ball with no brood and no queen — look harder. If you find a queen, suspect cold or starvation. If you find no queen after checking every cluster frame, consider late-season queen loss. A queenless colony stops rearing brood, the population ages out, and the remnant cannot generate heat.
  • A loose, sparse scattering of bees rather than a cluster — the colony collapsed before true winter. This is the varroa and viral collapse signature, and you should see supporting evidence in the brood.
  • Cluster wedged at the very top under the inner cover, frames below empty — they ate their way up and ran out of ceiling. Starvation in a hive that was simply too light in fall or ate faster than you planned for.

Count or estimate the dead. Spread the bees from the cluster onto your white tray. A dead cluster that fills a quart jar was roughly three to four thousand bees, which is far too small to survive a northern winter regardless of anything else. A cluster filling a gallon was ten thousand or more, and a cluster that size dying with food present means something acute: a prolonged cold snap combined with a short break in contact, or a disease event.

Step Three: Map the Stores Against the Cluster

With the cluster located, pull every frame in the brood boxes and sketch where the capped honey remains. Lay it out as a grid: box, frame position, roughly how much capped honey on each side. This takes ten minutes and tells you whether the colony could have reached its food.

Isolation starvation shows up as a dead cluster on one side of the box with two or three full frames of capped honey on the far side, or in the box above with nothing in between. Clusters move up more readily than sideways and will not cross empty comb in cold weather. If you see this pattern in multiple hives, the fix is not more feeding; it is better arrangement in October. Honey should be above and immediately flanking the cluster, with no empty frames between the bees and their food. Many beekeepers in cold regions move outside honey frames inward during the last fall inspection for exactly this reason.

Look at the honey itself. Crystallized stores are hard for a cold cluster to use. Fermented or sour-smelling stores, with uncapped cells and a yeasty odor, suggest high-moisture feed that was capped too late or honey from a wet fall. Dark, watery, bitter stores can be honeydew, which bees collect from aphid secretions and which causes dysentery when confined. Any of these can starve a cluster sitting on top of what looks like plenty of food.

Ready to put this into practice? Download HiveBook Free — it’s free and works offline.

Step Four: Read the Brood and the Comb

This is where varroa leaves its fingerprints. Pull the frames that held the last brood, usually at the center of the cluster, and examine the capped and uncapped cells under good light.

  • Perforated or chewed cappings — small holes in the center of capped brood, or cappings that look peeled back, show hygienic workers trying to remove mite-infested pupae. Scattered across a frame, this is strong varroa evidence.
  • White specks on the cell ceilings — tip the frame so you can see the upper inner wall of empty brood cells. Tiny white crystalline flecks are guanine, mite feces. Mites defecate on the cell ceiling, and even a handful of marked cells on a frame points to a heavy late-season infestation. Bee feces, by contrast, are brown and on the frame tops or comb face.
  • Dead pupae with deformed or missing wings — look at any bees that died while emerging or partially emerged. Shriveled wings mean deformed wing virus, which varroa transmits. Live-looking mites on dead pupae seal it.
  • Spotty brood with multiple eggs per cell or bullet-shaped drone cappings in worker cells — a failed queen or laying workers. The colony had no viable queen for weeks before it died.
  • Chalky white or gray mummies — chalkbrood. Rarely fatal alone, but a sign of a stressed, damp, or poorly ventilated colony.
  • Sunken, greasy, perforated cappings with a sour smell — stop and test. Insert a matchstick into a suspect cell and withdraw it slowly. If the contents rope out an inch or more like brown glue, assume American foulbrood until your state inspector says otherwise. Do not move any frame from this hive to another colony.

Also check comb condition. Comb that is black, heavy, and smells musty or ammoniated has been through many brood cycles and should be retired regardless of what killed the colony. Comb with chewed trenches and silk tunnels was hit by wax moth after death and is not evidence of the cause.

Guanine on the cell ceilings is the quietest and most reliable varroa evidence in a dead-out. Most beekeepers never look up inside the cell.

Step Five: What the Bottom Board Tells You

Finally, slide out the bottom board or screened board tray and look at the debris without disturbing it. The pattern of cappings on the board is a record of where the cluster was eating each week. A line of light wax cappings directly under the cluster is normal consumption. Cappings that stop abruptly show when the colony stopped uncapping food, which lets you estimate roughly when it died.

Now count mites. Take a flashlight and a hand lens and scan a few square inches of debris. Dead varroa are reddish-brown, oval, about the size of a pinhead, and shinier than the wax crumbs around them. In a colony that died of varroa, you will typically find dozens without trying. If you find none in a dead-out that otherwise looks like varroa, consider that the mites may have died and dried earlier in fall, or that robbing bees carried the last phoretic mites away. Also look for mouse droppings, small hive beetle larvae, and dead wax moth larvae, all of which arrived after the colony weakened rather than before.

Write down the mite estimate alongside your fall wash numbers. If your last August wash was under three mites per hundred bees and the bottom board is clean, varroa is unlikely. If the August count was six or higher, or you treated late and never rechecked, varroa is the leading suspect even if the cluster looks like it starved. Mite-damaged colonies starve easily because they are too small to generate heat and move to food.

Naming the Verdict

Most small-apiary dead-outs resolve to one of five causes once you combine the evidence:

  1. Varroa and virus collapse — small or scattered cluster, guanine in cells, perforated cappings, mites in debris, honey left behind, high or unchecked fall mite load. The most common verdict by a wide margin.
  2. Isolation starvation — cluster head-first in cells on one side, capped honey elsewhere in the hive, usually after a multi-day cold spell. Often in colonies that were also smaller than they should have been.
  3. General starvation — hive light, no capped stores anywhere, cluster at the top. A feeding or harvesting failure.
  4. Queen failure — no queen in cluster, drone brood in worker cells, or no brood at all going into fall. Often the colony was known to have superseded late.
  5. Disease or dysentery — heavy fecal streaking, foul odors, ropey brood, or chalk mummies, with the other patterns secondary.

Record the verdict with the evidence, not just the label. "Died by mid-January, two-frame cluster, guanine on three center frames, 40 lb honey left, last mite wash 7/100 on Aug 20, treated Sept 5" is a note you can act on. "Died over winter" is not. A per-hive history that carries these entries forward season to season makes the patterns obvious by the third year, which is exactly what HiveBook is built to hold, so the fall treatment decision is made against actual outcomes instead of memory. Livestock producers logging herd losses in Barnsbook and market gardeners tracking which beds failed in CropsBook are doing the same thing: the loss record is worth more than the loss.

If three hives in one yard all died the same way, the cause is your management, not bad luck. If one of five died, read that one carefully but do not rewrite the whole program around it.

Salvaging the Equipment Safely

Once you have your verdict, decide what to keep. Equipment from a colony that died of varroa, starvation, or queen failure is safe to reuse immediately. Shake out the dead bees, scrape the bottom board, and freeze any frames with pollen or brood remnants for 48 hours to kill wax moth eggs before storing them. Frames of capped honey from these hives are valuable: they are ideal for boosting a spring split or a new package, and far better than syrup. Mark them so you remember which hive they came from.

Comb that is heavily soiled with feces should be cut out and rendered rather than reused. Dysentery alone is not a reason to discard equipment, but if you suspect Nosema, a scorch of the woodenware and fresh foundation remove any doubt for a minimal cost. Any hive with suspected American foulbrood is a different matter. Do not reuse, sell, or give away those frames. Contact your state apiary inspector, and follow their direction, which in most jurisdictions means burning the frames and scorching or irradiating the boxes. Spores remain viable for decades.

Before you store the boxes, do one more thing: stand in front of the empty hive stand and ask what you would do differently with that colony in August and October. Treat two weeks earlier? Recheck after treatment? Combine it with a neighbor when it was clearly too small in September? Rearrange the honey frames? Write the answer in the hive's record where you will see it next summer. That sentence, read at the right time, is the entire return on the hour you just spent.

Dead-outs are part of beekeeping, and a loss rate of twenty to thirty percent in a hard winter is not a sign of a bad beekeeper. Repeating the same loss three winters in a row is. The autopsy is how you tell the difference. Work it slowly, write it down, and let the evidence, not the frustration, decide what changes in your apiary next season.