How to Count Varroa Mites in a Nucleus Colony: Adapting the Standard Method

A 100-bee sample from a 3-frame nuc provides only 55% confidence in threshold detection versus 95% for a 300-bee sample. That accuracy gap is real and significant -- but it doesn't mean you can't monitor nucleus colonies effectively. It means you interpret small-sample results with appropriate uncertainty and use VarroaVault's sample size calculator to understand what confidence level your sample actually provides.

TL;DR

A valid mite count sample requires approximately 300 bees from the brood nest for statistically reliable results
alcohol wash is 15-20% more accurate than sugar roll for detecting mite infestation levels
The calculation is: (mites counted / bees in sample) x 100 = infestation percentage
A 2% threshold triggers treatment in spring/summer; 1% is the fall action threshold
Count at least once per month during active season; increase to every 2 weeks if levels are near threshold
Log every count in VarroaVault to build a trend dataset that shows whether populations are rising or stable

Why Standard Sampling Is Difficult in Nucs

The standard alcohol wash protocol calls for 300 bees collected from the brood nest area. In a healthy full colony with 20,000+ workers, collecting 300 bees is easy -- you can scoop them with a measuring cup from a brood frame without meaningfully disturbing the colony.

In a 3-frame nucleus colony with 3,000-5,000 bees total, collecting 300 bees represents 6-10% of the entire population. That's a meaningful disruption to a small colony, and in practice it may be difficult to get 300 brood nest bees without taking such a large proportion that you stress the nuc.

The math challenge: for a standard 2% threshold, you'd expect to find 6 mites per 300 bees. In a 100-bee sample from the same colony, you'd expect 2 mites. The difference between 1 mite and 3 mites on 100 bees is the difference between "looks fine" and "needs treatment" -- and at that sample size, the statistical noise overwhelms the signal.

Adapting the Protocol for Small Colonies

Sample as Many Bees as Practical Without Stressing the Colony

For a 3-4 frame nuc, aim for at least 100 bees and ideally 200. Work from the brood nest area but stop if you've collected approximately 5% of the visible adult bee population. At 200 bees, your confidence in threshold detection is approximately 80% -- not as good as 300, but workable for routine monitoring.

Adjust Your Calculation

The standard formula still applies: mites found / bees sampled x 100 = infestation percentage. A 100-bee sample with 3 mites = 3% infestation. A 200-bee sample with 4 mites = 2% infestation.

Important: Do not simply scale your threshold based on sample size. If your actual infestation is 2% and you found 2 mites in 100 bees (which calculates correctly as 2%), that's still an above-threshold result at this time of year. The threshold doesn't change based on your sample size -- your confidence in the result changes.

Use a Lower Threshold for Action with Small Samples

Given the reduced accuracy of small samples, some beekeepers and extension programs recommend using a more conservative action threshold for small-sample results:

Standard alcohol wash on 300 bees: act at 2% in-season, 1% in fall
Small sample (100-200 bees): consider acting at 1.5% in-season, 0.8% in fall

This adjusts for the known underestimation risk at small sample sizes. If you found 1 mite in 100 bees (1%), you can't be confident whether the true infestation is 0.5% or 2% due to sampling variability. Acting at a lower threshold compensates for this uncertainty.

When to Recount vs. When to Treat

If you get a borderline result from a small nuc sample:

At 2x+ above your action threshold: Treat now. Even accounting for sampling variability, a result at 2x threshold is almost certainly above threshold in reality.
Between 1x and 2x your action threshold: Recount in 14-21 days with as large a sample as the colony allows. By then, the colony may be larger (allowing a better sample) or the trend will clarify the picture.
Below threshold: Monitor as scheduled, noting the reduced confidence in the result.

The VarroaVault Small Colony Sample Size Calculator

The small colony sample size calculator in VarroaVault adjusts the percentage calculation and confidence interval for samples under 200 bees. When you enter your mite count and sample size, the results screen shows:

Your calculated infestation percentage
The 95% confidence interval for your result (e.g., "likely between 1.2% and 4.8%")
The threshold recommendation adjusted for your sample size
Whether a recount is recommended given the confidence interval

This makes the uncertainty explicit rather than presenting a small-sample result with false precision. A 2% result from 100 bees looks the same as a 2% result from 300 bees in a raw number -- but the confidence intervals are very different, and the action recommendation should reflect that.

Alternative Methods for Very Small Colonies

For a 2-frame nuc with fewer than 2,000 bees, even 100-bee collection may be stressful. Two alternatives:

Brood cell inspection. Open 50-100 capped cells (worker brood preferably, not drone) and count mite-infested cells. The infestation rate you calculate from brood cell inspection doesn't directly compare to an alcohol wash percentage, but it gives you directional information about reproductive mite activity. The mite wash calculator guide explains how to interpret brood cell infestation rates.

Wait until the colony grows. If the nuc is very recently made (7-14 days old), wait until it reaches at least 3 frames of bees before the first count. A single count from a larger colony gives you more reliable information than multiple counts from an undersized one.

Monitoring Schedule for Nucs

The monitoring calendar for nucleus colonies differs from that for full colonies:

Newly created split or queenless nuc (0-21 days): Count at 14-21 days, during or just after the queenless period when most mites are phoretic. This is often the most accurate count window available for a new nuc.

Established nuc growing toward full colony: Count every 30 days as colony population allows, using the confidence interval adjustment for any sample under 200 bees.

Once nuc reaches 4+ frames: Standard monitoring protocol applies.

Frequently Asked Questions

How do I do an alcohol wash when I cannot find 300 bees?

Collect as many bees as practical from the brood nest area without exceeding about 5% of the visible adult population. At minimum, collect 50-100 bees for a directional result. The calculation method is the same (mites / bees x 100), but your result has lower statistical confidence than a standard sample. VarroaVault's small colony calculator shows the confidence interval for your actual sample size so you understand the uncertainty in your result. At 200 bees, you're at roughly 80% confidence in threshold detection; at 100 bees, approximately 55%.

How does a small sample size affect my mite count accuracy?

Small samples increase the role of random variation in your result. If the true infestation is 2%, a 300-bee sample would most likely return 4-8 mites (1.3-2.7%). A 100-bee sample might return 0-4 mites (0-4%) from the same colony due to random variation in which bees you happened to collect. This means small-sample results at borderline levels (near your action threshold) are less reliable decision drivers than large-sample results. VarroaVault shows you the confidence interval for your result, making this uncertainty explicit.

Does VarroaVault adjust my results for a small sample size?

Yes. When you enter your mite count and sample size, VarroaVault calculates both the infestation percentage and the 95% confidence interval for your result. The threshold alert system adjusts its interpretation for small samples -- for example, a result of 1.8% from 100 bees may trigger a near-threshold alert and recommend a recount, whereas the same result from 300 bees would be treated as a firm borderline reading. The results screen shows clearly when a recount is recommended due to sample size uncertainty.

How soon after treatment can I run a post-treatment mite count?

Wait 2-4 weeks after the treatment ends before running a post-treatment count. Counting too soon (within a week of treatment removal) may show mites still dying or emerging from the last brood cycle. Waiting 2-4 weeks allows emerging bees from brood that was capped during treatment to fully emerge and any surviving mites to become detectable in a new count.

What should I do if my mite count results seem unusually high or low?

If results seem surprising, repeat the count within 1-2 weeks before making a treatment decision based on a single outlier result. Confirm you sampled from the brood nest center (not outer frames), used the correct sample size (approximately 300 bees), and shook vigorously for the full 60 seconds. Consistent sampling technique is the most important factor in count accuracy.

Can I count mites from a sticky board instead of doing an alcohol wash?

Sticky board counts measure mite fall rate over 24-72 hours, which correlates with infestation level but is not a direct measure of infestation percentage. Sticky board results cannot be converted to an accurate percentage without calibration, and they are less reliable than alcohol wash for treatment decisions. Use sticky boards for general population monitoring but rely on alcohol wash counts for threshold decisions.

Sources

American Beekeeping Federation (ABF)
USDA ARS Bee Research Laboratory
Honey Bee Health Coalition
Penn State Extension Apiculture Program
Project Apis m.

Get Started with VarroaVault

An alcohol wash gives you the number. VarroaVault turns that number into a decision. Log your count, get an instant threshold comparison, and build a monitoring history that shows you whether mite levels are rising or stable across your entire operation. Start your free trial at varroavault.com.