• Mann Lake Ltd.

Thanks to these sponsors, you can enjoy this website without annoying popup ads!
You can show your appreciation by clicking on their banners above to go directly to their websites.

Print Friendly, PDF & Email

Simple Early Treatment of Nucs Against Varroa

First published in: American Bee Journal, April 2013

Simple Early Treatment of Nucs Against Varroa

Randy Oliver


First published in ABJ April 2013

Starting the season with a low level of varroa allows a colony to get a jump on the mite and its associated viruses.  I tested a simple method for incorporating varroa management into nuc production.


When I try to understand something about beekeeping, I seek out examples from the extreme ends of the spectrum.  For that reason, I often look to the experience of our Canadian brethren, due to their long, cold winters and bounteous honey crops.  We can take advantage of those huge honey yields to allow us to discern even small effects upon honey production from the impact of the varroa mite, eh?  Dr. Rob Currie found that surprisingly low mite levels can affect yield [1], recommending that the late spring varroa infestation rate not exceed 1 mite per 100 bees.  At this year’s ABF convention, Dr. Medhat Nasr described how beekeepers in Alberta, Canada find that they get best results with very early season mite control.

I’ve previously described how one can reduce mite levels by using queen cells to make up spring nucs [2]; but we can go a step further if we hit the nucs with a miticide at the same time!  The question then, is which miticides are gentle enough so as not to adversely affect the newly-mated queen or the buildup of the nuc?

Several studies have found that the synthetic miticides may have adverse effects upon queens, so I hesitate to use them.  That leaves the essential oils and organic acids.  I know from experience that the most effective essential oil—thymol–is disruptive to broodrearing, so I’d rather not apply it to small nucleus colonies (there are also temperature issues with thymol).

As far as formic acid is concerned, the manufacturer of Mite-Away Quick Strips™ recommends that they be applied only to colonies exceeding 6 frames in strength [3], and anecdotal reports from a number of beekeepers suggest that formic acid, under some circumstances, may be risky to queens–an observation supported by research by Dr. Pierre Giovenazzo [4].

So that leaves us with only two proven miticides–the recently-registered product HopGuard® (hops beta acids) and the unregistered oxalic acid.  The manufacturer of Hopguard states that the product is safe for queens [5], so it sounded promising.  The oxalic acid dribble was a likely candidate as well, since it also does not appear to negatively affect queens [6,7].  Giovenazzo [8] had also tested oxalic acid with good results.

However, both Hopguard and oxalic have one drawback—since they only kill phoretic (hitchhiking) mites, and since either product is only active for a few days in the hive, they don’t hit the reservoir of mites in the brood.  Both of these miticides are most effective in broodless colonies, such as in fall, or during a period of induced broodlessness, as demonstrated by Wagnitz and Ellis [9]–who caged the queen in late summer, replaced her in a few days with a queen cell, and then later applied oxalic acid after all the brood had emerged.

I normally start nucs with queen cells.  It occurred to me that in such nucs, a window of opportunity exists for the effective use of short-term natural treatments against varroa.  It’s all about the timing.  A nuc is made up with frames of brood from an established colony.  That brood will contain mites.  Some of the unsealed brood will continue to be invaded by mites for up to 9 days after the nuc is made up (bottom colored bar).  But any and all brood from the parent queen will have emerged by Day 21 after the making of the nuc (Fig. 1).

Figure 1.  The theory behind the early treatment of nucs—it’s all about timing!  There is a brief window of opportunity from Day 19 to Day 21 after make up in which every mite in the nuc is forced out of the safety of the sealed brood.  A short-term treatment applied at that precise time could result in a very effective kill of the now-exposed mites!

I insert 10-day* queen cells into the nucs on the day after I make them up.  That means that the new queen won’t emerge until Day 2 or 3 after make up (middle colored bar), and not begin laying eggs until around Day 11 after make up (sometimes a bit sooner).

But the mites cannot yet enter the new brood, since varroa doesn’t invade a cell until about 8 days after the egg is laid [10].  That means that the first opportunity for the mite to hide in new brood generally occurs around Day 19 post make up of the nuc (upper colored bar).  So from Day 19 through Day 21, virtually every mite would be exposed to the treatment!

OK, this sounds good in theory.  So I ran two trials to see just how well it actually worked in practice.

* I no longer use cells any more ripe than that, because since I started selecting for mite resistance, the occasional batch of cells will emerge on Day 11, rather than on Day 12 after grafting.

Trial #1

Materials and Methods

The unusually warm winter of 2011-2012 was a good opportunity to test the method, since mite levels were unacceptably high by early April.  We used a batch of nucs grafted from two queen mothers (the majority from one mother) on April 12.  On May 1 (Day 19 after make up) we equalized them to 48 queenright nucs each containing 5 full frames of bees by adding frames from the unmated nucs to the mated ones, and by shaking bees, which helped to randomize the original mite infestation rates.  At this point (again apparently due to warm weather), some larvae from the new queens were already being sealed, meaning that some mites may have already infested those cells prior to treatment.

Since the nucs were scatted in a rough line in the order of make up, every 4-5 in the row would have come from the same parent colony, so we marked them sequentially down the line for treatments in order to avoid any effect from the original brood sources.  After allowing 2 hours for them to settle down, we took samples of ½ cup of bees (~320 bees) from each nuc, preserved them in alcohol for later washing for mites, and then applied treatments as below (Table 1).

Treatment Application
Control Open and smoke alone
Hopguard® 1 strip in center of nuc
Hive Clean®* 1- 15mL packet dribbled evenly over seams of bees
Oxalic acid dribble 5mL per seam of bees  3.2% w:v oxalic acid solution in 1:1 syrup**
* Bee-Vital Hive Clean® is a widely-used product from Austria, containing water, sugar, oxalic acid, citric acid, formic acid, and propolis.


Table 1. Treatments used on nucs, 12 colonies in each group.

After a week, we moved the nucs to another yard, placing them in groups of 4 facing out 90 degrees to each other, and rotated to equalize the directions of the entrances for the various treatments.  Shortly afterward, we worked each nuc into a single, adding 5 frames of foundation.  We fed 1:1 sucrose syrup equally as necessary to augment the natural honey flow.

We took mite samples again at Day 37, Day 51, and Day 87 post treatment.

Results and Discussion

The mite infestation rates of the groups are shown in Figure 2 (Day 0 is now reset from the treatment date).

Figure 2.  Changes in mite infestation rates over 87 days (approximately 5 to 6 mite reproductive cycles). Note that the infestation rates at Day 0 were exaggeratedly high due to there being no brood in which mites could hide.  Standard errors of the means indicated.

The differences between the first two bars are the most indicative of the efficacy of the treatments, with the greatest reduction being from the oxalic treatment.  Mite infestation rates climbed at a fairly steady rate after treatments.  My treatment threshold is 2 mites per 100 bees.  This level was exceeded in the control group (which began with the lowest mite level) by the first time point.  By contrast, the mite level in the oxalic group (which began at nearly twice the mite infestation rate of the controls) was still well below threshold at three months!

To more easily compare the effects of treatment, I normalized the mite population growth curves for all groups to start at 100% (Fig. 3).

Figure 3.  Normalized curves of mite population growth.  The mite infestation rate nearly tripled in the control group over the course of the trial.  The mite rate of the oxalic group even three months after treatment was only slightly higher than half the starting rate!

The mite count data need to be taken with a bit of caution, as they were only single samples from each colony at each time point, and thus have a built in degree of potential error, especially in the low ranges, which give disproportionate influence to any single mite in (or not in) the sample.  However, I’ve carefully inspected the raw data, and feel that the results are meaningful, despite the variability in counts.

The intermediate performance of Hopguard and Hive Clean (applied at manufacturer’s recommended rates) suggests that their efficacy was less than that of the oxalic dribble.  The registrants may need to adjust the suggested treatment rate for nucs.

Note also that each colony had received a new queen, who may have passed on mite resistance to her offspring.  Indeed, in 3 of the 10 colonies in the control group which made it to the end of the trial, the mite counts were lower at the end of the trial than they were in the beginning.  But compare this to the oxalic group, in which mite counts went down in 9 out of 11!

Colony Survival and Productivity

We removed 9 colonies during the course of the trial due to failure or disease (EFB), roughly spread among groups.  The oxalic group had the lowest rate of failure, with only one removal.

Measuring the productivity of the nucs was problematic, since the main honey flow essentially failed.  On July 27 (Day 87 post treatment) I opened every hive in the test apiary, excluding (censuring) any that had superseded, or with abnormally small populations.  I recorded which colonies fell into one of two extremes from the norm (which had roughly filled a third to a half of the second deep with honey)—as “productive” (having nearly filled the second deep with honey) or as “nonproductive” (having barely touched the foundation).  The results were unexpected:

  • · Of the 13 “productive” colonies, only 3 had received an acid treatment (oxalic or Hive Clean).
  • · Of the 7 “nonproductive” colonies, 6 had received some form of acid treatment.
  • · The Hopguard group contained the highest proportion (6 of 10) of productive colonies
  • · Of the 13 “productive” colonies, 4 had high mite counts (4.7-13 mites/100 bees).

I don’t know whether the apparent lack of production of the acid-treated colonies was a fluke, or whether the acid treatment had some sort of long-term effect upon productivity–the lack of normal honey flow may have confounded the results.  Giovenazzo [11] also observed a nonsignificant 13% reduction in honey yield after oxalic treatment, but he applied twice the dosage of oxalic acid as I did.    This potential effect certainly demands further investigation!  On the other hand, compare this result to the grading for colony strength in Trial 2.

Trial #2

Materials and Methods

We ran a second trial with oxalic dribble alone to see whether we would obtain similar results as from Trial 1.  Grafting (all from the same queen mother) took place May 3, and we made up 4-frame nucs 9 days later.  In this trial, the weather was warm, and the queens started laying unusually early, with mature larvae at Day 15 after nuc make up.  We equalized 36 queenright colonies to 5 frames of bees on that date.

We alternately treated the hives the next day (May 28), with either oxalic dribble or sham opening, but did not take initial mite counts.  This treatment timing was earlier than optimum, since workers from the parent queen would still be emerging for 5 more days after treatment, possibly compromising the efficacy of the treatment.

After a week, we moved the hives to another yard and worked them into singles, adding 5 frames of foundation.  The honeyflow failed to materialize in June (but pollen was abundant), so we fed the colonies equally with 1:1 sucrose syrup.  The strongest colonies were just filling the 10th frame at grading on Day 69 after making the nucs (Fig. 4).

Figure 4.  Timeline of Trial #2.Results and Discussion

All colonies in the oxalic group survived to the end of the trial; three failed in the control group.  Again, the oxalic dribble substantially suppressed mite infestation rates.  I present the results differently here, showing the distribution of mite rates across the treatment groups (Fig. 5).  The green bars represent the control group, which had a median value of 4 mites per 100 bees at Day 53 post treatment, compared to a median of 1 mite per 100 bees for the oxalic-dribbled group.

Figure 5.  The mite count of the control group was distributed around 4/100 bees, with 2 colonies having excessive counts.  On the other hand, 12 of the 18 OA-treated colonies had counts of 1/100 bees or less.

I observed no negative effects due to treatment of the nucs with oxalic dribble.  Only two colonies in the entire yard went queenless—both were in the untreated group.  Overall, the oxalic-dribbled colonies were substantially stronger at 53 days (2.5 brood cycles) after treatment (Fig. 6).  This result reflects those of Giovenazzo [12], who also observed stronger colonies after oxalic treatment (although not statistically significant).  One plausible explanation for this result is that the knockback of mites just prior to the first round of brood being sealed is enough to break the virus infection cycle of the first generation of bees, allowing for greater longevity of those bees.

Figure 6.  Distribution of colony strength at the end of the trial.  The median strength of the control group was 7 seams of bees; of the oxalic-dribbled group, 9 seams.  This result suggests that any negative effect of the oxalic dribble was more than compensated for by the benefit of mite reduction.

I was curious as to whether differences in strength of the colonies was related to nosema infection, so I checked a 20 house-bee sample from each of the three weakest, and three strongest colonies, with representatives included from each treatment group.  None of the strongest colonies showed nosema spores, but two of the weakest did—one of which showed 30 spores per field of view (1 mL/bee dilution).  I squashed an additional 10-bee sample from that colony one bee at a time—only 1 of the 10 was moderately infected.

So, did colony strength reflect the mite infestation rate?  I plotted colony strength vs. final mite count (Fig. 7).

Figure 7.  There was a distinct trend that those colonies with higher mite counts tended to be weaker, although the correlation was weak, perhaps due to the lack of precision in single-sample mite washes.

Overall Discussion

Despite the fact that in both trials some brood had already been sealed by the time I applied treatment, the method was not only very effective at reducing mite levels (to 1 per 100 bees in most colonies), but also inexpensive (pennies) and quick.

Based upon the early results, we treated several hundred nucs this spring with oxalic dribble at Day 19, and did not notice any difference in queen failure over our normal low rate.

Practical application:  following only two mite treatments in the past 9 months (one oxalic dribble in November, and the oxalic dribble over the nucs in May)  our mite counts across the board in late July were still gratifyingly low—averaging a bit less than 2 mites per 100 bees (some of this was also due to breeding for resistant stock).

But how in the world, you say, will I be able to keep track of treatment window dates during the hectic spring season?  That was also a major concern to me, since during our spring nuc making frenzy I often wouldn’t be able to tell you the day of the week!  I solved the problem by printing up a simple spreadsheet (Fig. 8) that I could check each morning.  For the cells grafted on any day, it shows the two critical dates in red—the last day that we can make nucs for that batch, and the 19th day for queen check and treatment.

Table X.  A portion of my queen rearing spreadsheet for 2012, which helped me to keep my dates straight.  The two critical dates are in red.  In order to avoid weekend commitment as much as possible, I don’t graft on Wednesdays or Thursdays.  I fill in the grafting details and yard in which the nucs are placed.

The above spreadsheet made it really easy to pull off the timing of treatments despite my perpetual disorganization (and actually made me feel somewhat professional)!  The method only required one slight change in our regular production of nucs.  We normally check for queen rightness two weeks after putting in the cells (good mating weather permitting), but in order to save trips to the nuc yards, we now wait 19 days, so that we can do three things on the same visit:

  1. Check for laying queens.  On Day 19, any queens from the grafted cells should normally have a good pattern of open brood, and it is just before the date that any emergency queens or laying workers would have started laying.
  2. We combine the frames of bees from the unmated nucs with the queenright ones to boost them all to 5 frames of bees.
  3. We then dribble them with oxalic acid before putting the lids back on.

By this method, the added oxalic dribble only adds a few seconds per nuc to our normal routine, plus by waiting a few more days to check for laying queens, we weed out the early failures or poor layers.

Possible Improvements On The Method

In warm weather, there may be brood from the new queens being sealed a few days earlier than Day 19, so you should check to see whether you need to treat earlier.  If you find this to be the case, it may be of benefit to make the nucs up a few days before the cells are ripe, to allow the original brood (and mites) enough time to emerge.

The efficacy could also be improved by treating the parent colonies of the nucs with formic acid a few days prior to splitting them.  If one makes up nucs by the “yard trashing” method (the complete breakdown of the parent colonies into nucs), any queen loss due to the formic treatment would make little difference.

Since mites continue to enter brood in a nuc for 8 days after make up, efficacy could potentially be improved by applying an additional oxalic dribble or Hopguard strip at make up.  Let me also make clear that I have not given up on Hopguard or Hive Alive (should it be registered in the U.S.)—both show potential.

Updates August 2022:  If the old queen is still “looking good,” we return her to a single brood chamber placed back on the parent stand.   We stock the single only with combs containing minimal sealed brood (so that few mites are protected), along with some nurse bees.  The field force quickly returns to cover the combs, and once they do, we dribble the colony with oxalic acid, killing nearly all the mites.  The queen, who is in full laying condition, quickly fills the combs with brood, and the colony quickly recovers its strength.  We replace the then-aged queen later in the season.

Once the nucs are mated and dribbled, let the queens lay for a while.  You can then recombine any of them to create strong, nearly mite-free colonies with fresh queens.

Also refer to:  MITE CONTROL USING A 14-DAY BROOD BREAK WITH OXALIC DRIBBLE at First year beekeeping – Scientific Beekeeping.

And if you simply want to split a colony in two, try this:

Bottom line:  These methods of creating an induced brood break allows for very inexpensive and reliable mite control.  It has allowed us to avoid using synthetic miticides for over 20 years.  I don’t make recommendations, but the mite-control strategy below has worked well for us for many years.


This method uses precise timing, combined with making normal colony increase, to gain the most advantage of residue-free “natural” mite treatments.  The oxalic dribble costs pennies and takes seconds to apply.  We already love it for early winter treatment at cessation of broodrearing, and now can also use it in spring.  Our findings also call for more research on the possible effect of oxalic dribble on productivity, and whether treatment with two Hopguard strips would give better results.

Practical consideration:  this project was funded by donations from beekeepers, performed by beekeepers, for the benefit of beekeepers.  You can support such research with your donations to ScientificBeekeeping.com.


I greatly appreciate the help in running this trial from my sons Eric and Ian, whose labor was covered by your generous donations to ScientificBeekeeping.com. I especially wish to thank volunteer Brion Dunbar for his unstinting assistance throughout the trial.  The Hive Clean was generously donated by BeeVital, Seeham, Austria.


[1] Currie, RW and P Gatien (2006) Timing acaricide treatments to prevent Varroa destructor (Acari: Varroidae) from causing economic damage to honey bee colonies. Can. Entomol. 138: 238–252.

[3] (Broken Link!) http://www.miteaway.com/uploads/3/0/7/9/3079637/_maqs_application_brochure.pdf

[4] Giovenazzo, P and P Dubreuil (2011) Evaluation of spring organic treatments against Varroa destructor (Acari: Varroidae) in honey bee Apis mellifera (Hymenoptera: Apidae) colonies in eastern Canada.  Experimental and Applied Acarology 55(1 ): 65-76.

[6] Cornelissen, B, et al (2012) Queen survival and oxalic acid residues in sugar stores after summer application against Varroa destructor in honey bees (Apis mellifera).  Journal of Apicultural Research 51(3): 271-276.

[7] Wagnitz, JJ and MD Ellis (2010) The effect of oxalic acid on honey bee queens. Science of Bee Culture 2(2) (Supplement to Bee Culture magazine 138(12): 8-11. http://www.beeculture.com/content/ScienceJournalDec2010.pdf

[8] Giovenazzo (2011) Op. cit.

[9] Wagnitz, JJ and MD Ellis (2010) Combining an artificial break in brood rearing with oxalic acid treatment to reduce varroa mite levels. Science of Bee Culture 2(2) (Supplement to Bee Culture magazine 138(12): 8-11. http://www.beeculture.com/content/ScienceJournalDec2010.pdf

[11] Giovenazzo (2011) Op. cit.

[12] Giovenazzo (2011) Op. cit.