If it's late summer or autumn as you're reading this, this is the window that matters. Mite numbers peak while the colony contracts, and the bees being raised now are the ones meant to carry the colony through winter. Damage done in September shows up as an empty box in December.
Varroa is the reason most backyard colonies die in the United States. Not directly, mostly: the mites carry and amplify viruses, and it's the viruses that hollow a colony out. Which is why a hive can look busy in August and be gone by November with its honey untouched.
The good news is that it's measurable and manageable. The bad news is that both parts get skipped, and "I treated in the autumn" without a count either side is close to guessing.
Why a real count, not a guess
You cannot tell a colony's mite load by looking at it. Bees don't look ill, the colony doesn't act sick, and by the time you can see deformed wings the problem is well advanced. Seeing a few mites on bees tells you very little; seeing none tells you almost nothing at all.
A count is a sample of around 300 bees taken off a brood frame, washed, and the mites counted against the bees. It gives a percentage you can track over time, compare against a threshold, and use to decide something.
The threshold most people work to is around 3 mites per 100 bees, lower going into autumn. What matters more than the exact number is that you have one at all, and that you took it the same way each time.
The methods, honestly compared
| Method | What it's good for | The catch |
|---|---|---|
| Alcohol wash | The most reliable number. Consistent, repeatable, hard to do badly. | Kills the ~300 bees in the sample. |
| Sugar roll | The bees survive. Fine if you're squeamish about the wash. | Tends to under-report, and humidity here makes it less consistent than it looks. |
| Sticky board / natural drop | Non-invasive, no sample taken, useful for a rough trend. | Not directly comparable to a wash figure, and easily thrown off by weather and colony size. |
I use alcohol washes, because the number is worth having. Three hundred bees is a very small fraction of a colony, and it buys a figure you can actually act on.
Applied is not the same as delivered
This is the part that catches experienced beekeepers, not just beginners. A treatment can be correctly bought, correctly placed, and still deliver almost nothing:
- Strips get chewed out. Some colonies are ferocious housekeepers and will remove a strip long before it's finished working. You find shreds on the bottom board and assume the colony was treated.
- Placement matters. A strip away from the cluster is a strip doing nothing.
- Temperature limits are real. Several products have a window, and outside it they either don't work or hurt the colony. Our coastal swings are not always kind to that.
- Mites under cappings are out of reach of most treatments, so timing against the brood cycle changes the result entirely.
- Fed products depend on the colony taking the feed, which it may not during a flow.
The only way to know is to count afterwards. Test before, treat, test after. If the number didn't move, you learned something important while there was still time to act.
Choosing a treatment
There isn't one right answer, and anyone who gives you one without asking questions isn't paying attention. What it depends on:
- Whether there's honey on the hive, which rules several options straight out
- Temperature over the treatment window, not just today
- Whether the colony has brood, and how much
- What you used last time, since rotating matters for resistance
- How you like to manage, because a treatment you're uncomfortable applying is one that gets applied badly or late
Whatever we land on, it gets used exactly as the label says. These are regulated products, the label is the legal instruction, and most of the ways treatments fail are ways the label already warned about.
Non-chemical controls are part of it too
Chemistry isn't the whole toolkit. Drone brood removal, brood breaks, and requeening with mite-resistant stock all reduce mite pressure, and they work best planned into the season rather than reached for in a crisis. Worth talking about in spring, not October.
What I actually do
I come out, take a proper wash from a brood frame, and give you the number there and then. We talk about what it means for this colony at this time of year, choose a treatment together if one's warranted, and set a date to count again.
You'll also learn to do the wash yourself if you want to, and I'd rather you did. It's a fifteen-minute skill that changes how you keep bees.
If you've already lost a colony this year, reading what happened is the place to start, because the answer usually shapes what the survivors need.
Frequently asked questions
How often should I test?
At least monthly through the season, and always before and after a treatment. Around every three weeks is roughly one worker brood cycle, which gives a clean read on which way the population is moving.
Does the alcohol wash kill the bees?
The sample, yes, about 300 of them. That's a very small share of a colony and it buys a number you can trust. Sugar rolls spare the sample but tend to under-report.
I treated. Why are my counts still high?
Applied isn't delivered. Chewed-out strips, poor placement, wrong temperatures, and mites sealed under cappings all produce a treated colony with an untreated mite population. A count afterwards is the only way to catch it.
My hive looks strong. Do I really need to test?
Yes, and that's exactly the trap. A strong colony in July can be hollow by November, because the damage lands on the winter bees before it shows.
Can I just not treat?
Some people go treatment-free deliberately, with resistant stock and a plan and an acceptance of higher losses. That's a legitimate approach. Not treating because you haven't got round to it is a different thing, and it usually ends the same way.
Is it too late in the season to do anything?
Depends what the count says and what's in the box. Late is worse than early and much better than never. Get a number first.
What areas do you serve?
Santa Barbara, Goleta, Montecito, Carpinteria, Isla Vista and the surrounding South Coast.