• Dadant
  • Mann Lake Ltd.
  • Shastina Bee Girls

I'd like to thank these sponsors for supporting this website. Just click on their ads to go to their websites.

IPM 5.5 Fighting Varroa 5.5: Biotechnical Tactics II

Save pagePDF pageEmail pagePrint page

Introduction 1
Drone Brood Management and Trap Combs 1
Powdered sugar dusting 4
The Oliver 15-second sugar dust method 6
Discussion 7
The one-two punch—30 seconds to knock out varroa! 7
My new website 8
References 8

Tactics: Biotechnical Methods II–The one-two punch
Sixth in a series on Integrated Pest Management of varroa

Note: this article is broken into two web pages. Powdered sugar dusting details are on the next page!


None of the biotechnical methods that I detailed last month will control varroa alone. Now it’s time to give you some real meat! Either of the two methods I’m about to describe, used alone, have been tested and proven to keep varroa at tolerable levels if performed properly. Used together, they may be a nonchemical one-two punch that will give us the upper hand on the mite.

A note to those commercial beekeepers who are saying, “Jeez, this guy is totally whacked out! I’m already maxed out for time, and can’t be putting special labor-intensive gizmos into my hives—that’s for hobbyists. Plus, I can’t take a chance on trusting my mortgage on some untested mite control method.” In answer, my California son would say, “I feel you, man.” I’ve got two groups of beekeeping friends—hobbyists who are willing to test new things on their handful of colonies, and commercial guys who run very efficient operations by rotating a series of ag chems. What I’ve been seeing the past few years, are hobbyist/sideliners maintaining strong colonies without any synthetic miticides. I was skeptical as hell at first, but they’ve demonstrated the feasibility. As with any historical change in an industry, the majority says, “It can’t be done, we can’t afford it, it won’t work in reality.” Then, several years later, they look back and see that it was done, the industry could afford it, and it did work in reality.”

Drone Brood Management and Trap Combs

The first punch that we’re going to hit the mite with is based upon the fact that varroa reproduces rather poorly in worker brood, but is nearly three times more successful in drone brood, due to its longer postcapping period. It’s not surprising then, that female mites prefer drone brood by a factor of roughly 10 to 1 (reported figures range from 4:1 – 12:1). The mites, being tiny and blind, apparently recognize nurse bees by odor (Dillier 2004), and ride around on them until they smell a drone larva of the right age. Since nurse bees spend much more time feeding drone larvae than worker larvae (Calderone & Kuenen 2003), the mites have ample opportunity to come into contact with drone larvae.

A feral colony of bees builds about 17% drone comb (Seely 2002). Rapid mite reproduction in this amount of drone brood largely accounts for the decimation of the feral bees by the mite. The stimulus to build drone combs is good forage (at any time of the year), with a negative feedback from drone brood already existing (Charriere, et al. 2003). Beekeepers, by using worker-sized foundation, can typically keep drone cells down to about 4% if they regularly cull old combs. However, colonies will normally produce temporary drone cells in the space between the brood chambers in spring. Indeed, a quick inspection of the exposed drone brood when you break the brood chambers apart can give you an indication of varroa infestation level.

The beekeeper practicing varroa IPM can minimize varroa reproduction by managing the amount of drone comb in his colonies. This is especially important since hygienic bees remove only infested worker pupae, not drone pupae. I’ve already mentioned the importance of culling old combs with drone cells. Wilkinson and Smith (2000, 2001) modeled the effects of drone brood management. They state: “At 5% drone brood, as many mites are emerging from 50-60 drone cells as from 1000 worker cells. This certainly emphasizes the importance of drone brood in mite population growth, and the need for beekeepers to prevent large quantities of drone brood being reared unnecessarily and being left to emerge in the hive.” They suggest “regular and ruthless ‘culling’ of the old combs and the badly built combs.” Their model predicted that reducing drone brood from 4% to 3.2% would reduce the mite population growth rate by 25%! They suggest that drone brood is more important to mite growth at low mite levels, since drone brood capacity for mites reaches its limit before that of worker brood.

Clearly, the beekeeper should cull frames containing drone comb. However, we can go even a step further, and use drone comb to “trap” mites, and then remove those mites from the colony. This process is called “drone comb trapping,” and is widely used with great success in other parts of the world. The concept is simple: insert a frame of drone comb into a colony at the edge of the brood nest, allow the queen to fill it with drone eggs, wait while the mites infest the cells, then remove the frame before the mites emerge. Theoretically (Wilkinson & Smith 2002), trapping with one deep drone frame once a month for four months will delay the mite population from reaching a damaging level for 2-4 months; two frames monthly will delay it for a year.

So, you ask, theory is fine, but how effective is drone trapping in real life beeyards? The short answer is, surprisingly effective! Dr. Nick Calderone has an excellent guide at http://www.three-peaks.net/PDF/Drone%20Brood%20Removal%20for%20the%20Management%20of%20Varroa%20destructor.pdf. In his study (Calderone 2005), two combs were replaced monthly from June through September. Mite levels were kept to about 2.5% (ranging from 0-7%)—up to 10 times less than control colonies! The drone-trapped colonies also made more honey!

In Europe, Charriere, et al. (2003) report that drone trapping has no negative effect on the development of the colony and honey production. In their tests, which used the equivalent of one drone frame per colony, removed regularly, they found varroa buildup was suppressed enough that only a fall treatment with a natural chemical was required.

Drone trap combs clearly work, but are they practicable? For the hobbyist, the green plastic drone combs available from several bee catalogs are great. You put them in, wait exactly four weeks (a few days until the queen can lay, then 24 days for the drone development period) and remove them. The brood and mites can be killed by removal with a cappings fork, scraping, freezing, heating, or treating with formic acid. The combs are then replaced for another cycle. Reusing drone comb may have the added benefit, that it might be more attractive to mites. In an experiment where old combs were placed side by side with new combs in colonies of Brazilian AHBs, Piccirillo and DeJong (2004) found that cells of old combs were four to five times more attractive to varroa than same-sized cells on new combs. The authors concluded, “It is clear that these mites strongly preferred old worker brood comb cells to new worker brood cells.” Drone cells were not tested, but this avenue calls for further study.


One doesn’t need to purchase plastic trap combs. A medium frame, a deep frame with a section of comb cut out, or even a foundationless frame, will allow bees to produce volunteer drone comb, which can be cut out and discarded (or melted for the wax).

For the commercial beekeeper, it’s obvious that an extra piece of equipment, or the concept of freezing combs is impractical. Also, reaching down to remove broken-off pieces of drone comb built on the bottoms of frames would be too time consuming. To address those issues, I designed and tested a dedicated trap comb that can remain in the hive all year. I knew that the bees would store honey at the top of a comb, so I made a provision for that. I also wanted a wooden rim all around the drone brood so that I could quickly cut the comb out with a hive tool in the field.

I also knew that mites only enter drone cells on days 8 and 9. Therefore, each cell only has a two-day trapping window. Once a cell is capped, it can no longer trap mites. Ideally then, to ensure continuous trapping, one would want the queen to lay eggs progressively on the trap frame from the time it is inserted, until 9 days before it is removed. Therefore, I wanted to force the queen to lay progressively by making the bees build their own drone comb from scratch.

See the photo for the design of the Oliver Trap Frame. I simply take ordinary deep frames, and an extra top bar. I cut the ends off the top bar and install it upside down, slipping it onto a piece of plastic worker foundation ripped down to 2½” (Permadent® fits in the grooves better than deep cell foundation). There is a little over 2” of foundation remaining exposed at the top of the frame. This design works great! We tested 300 of them last year. Virtually every colony builds them out as illustrated—honey in the top, drone brood below (do not increase the 2” foundation dimension, or some colonies will produce worker brood above. We may wish to even decrease this dimension—let me know if you try).

Oliver drone trap frame

The Oliver drone trap frame.

Stephanie loading trap frames

Stephanie loading new trap frames. We tested 300 last year.

Inserting a trap frame

Inserting a trap frame in the almonds for an early start.

Drone trap frame at 4 weeks

Drone trap frame at four weeks—honey at top, drone brood below.

Cutting out the drone comb with a hive tool

Cutting out the drone comb with a hive tool.

Reinserting the frame

Reinserting the frame for the next round of trapping. The bees will quickly build new comb on the remaining wax. The whole process takes only 15 seconds!

Trap frames set to edge of brood chamber

Trap frames set to edge of brood chamber for comb honey production in bottom half.

Pure beeswax

Pure beeswax extracted from cut out drone combs. Less than an ounce per comb, but this wax can be sold at a premium, as it is completely pure and chemical free.

Here are the advantages of this design:

1. It takes only about 15 seconds per colony to open the lid, remove the comb, cut out the drone comb with your hive tool, replace the frame, and close the lid. It’s so fast that we don’t even close the door to the truck when we hit a yard! No freezing or extra work is required.

2. Since the bees must build comb from scratch, the queen can only lay so many drone eggs per day. This restraint extends the period that the combs are actually trapping mites.

3. Since the combs are returned to the same hive, there is no spread of disease from colony to colony.

4. When you are done with drone trapping for the season, move the comb to the outside of the cluster to produce comb honey for sale or winter stores.

We found that colonies with drone trap frames tended not to produce drone comb between the boxes. This observation is supported by Seeley (2002), who found that colonies with added drone comb build 7½ times less volunteer drone cells as those provided with 20% drone comb. So by adding drone comb, you actually remove the incentive for the bees to produce drone cells elsewhere in the colonies. In effect, drone trap frames allow you to manage drone production in your operation. Indeed, in our queen rearing operation, by removing the unwanted drones from poor queens, and by leaving extra drone combs in our drone mother colonies, we produce an excess of genetically superior drones for mating, while suppressing the population of genetically undesirable drones. Since we breed for mite tolerance, our drone mothers have fewer mites to start with, and the extra drone comb is less of an issue for mite buildup.

You may ask whether it is worthwhile to extract the wax from the cut out drone comb. We boiled 200 cutouts, and produced 10 lbs of wax. We found that it wasn’t worth our labor, so we now just feed some drone brood to the chickens, and compost the rest for the garden. If you had a better means of extracting the wax, you might be able to reclaim it. Or, you might find a market for it as livestock or pet feed, or sell it in Asian markets as a delicacy!

As I mentioned in the previous installment, using these combs in conjunction with queen restriction can nearly completely eliminate varroa from a colony (Calis, et al. 1999).

Zachary Huang is developing the Mitezapper—a drone trap frame with heating wires in the foundation. Once a month, the beekeeper would hook up wires to a car battery for a few minutes to kill the mites with heat. The colony would not even need to be opened!

Hoopingarner (2001) does raise one potential objection to trapping with drone brood: “it exerts constant selection against the mites that prefer drone brood. This is not in the long-term best interest of a varroa reduction program” because, it de facto selects for mites that prefer worker brood. However, upon further reflection, Charriere (2003) states, “The often expressed fear that removal of drone brood will select for a population of varroa that prefer worker brood does not seem to be justified. We should remember that the removal of drone brood occurs only during a short period, and for the rest of the year the mites are obliged to breed in worker cells.” Indeed, if we breed for varroa sensitive hygiene, the mites don’t stand a chance in worker brood.

Make sure that you read the updates to the end!

Updates 12/6/2010

French study: Efficacy of drone trapping four times in a season. Kept mite levels substantially (25%) lower, but still needed additional control measures. Honey production did not suffer as a result of trapping. Unexpected result was control of swarming!


California beekeeper Jeremy Rose reports: I did a few drone comb removal cycles on 130 hives this past spring, before the honey supers became too heavy to lift. I don’t know that I noticed reduced swarming this year (30% of my hives swarmed even though they had been split!), but I did catch the hives that wanted to swarm before the swarms actually left (helpful so that I knew to get them a new queen the following week). It was easy to see, since the drone frames would always have swarm cells on them. Hives that were not getting ready to swarm would have some cells of fresh eggs in the drone comb which made non-swarming hives easy to distinguish (I took partially-full supers off hives that were preparing to swarm and gave them to the non-swarming hives).

Bottom line:

Drone brood trapping works great, and can be done very quickly and cheaply. It doesn’t decrease honey production, and keeps the bees from building volunteer drone cells elsewhere. It may keep mites below economic injury levels alone, but will likely require supplemental treatments.

Points to remember:

1. A full comb removed monthly will generally keep mite levels below threshold.

2. Two full combs would be even better.

3. Two combs, alternately removed every other week, would likely be best.

4. Do not forget to remove the combs at 4 weeks, or you’ll be breeding mites!

Update 12/25/08

Drone trap frames left unmanaged in position 4 in the upper brood chamber in my California/Nevada bee management are used heavily in spring and early summer. Later in the season, the bees may fill them with honey, or they may make a mix of drone brood of uneven age with honey interspersed. Surprisingly, I find that in those colonies in which I accidentally left the drone frame in the broodnest (instead of moving it to the side) didn’t appear to have much different mite levels than those adjacent colonies whose frames I had moved.

I’m trying to get a feel for the value of drone trap frames. The one property of them that stands out to me is that they can be used to “manage” where the bees place drone comb in the colony, by giving them a place to build it.

It may be that a partial frame of drone brood, as in my trap frames, may satisfy the colony’s desire to build drone comb and raise drones, thus eliminating most volunteer drone brood elsewhere. It may also be that we might be able to strongly affect varroa buildup with only one or two removals of the patch of drone brood at an appropriate time.

The question would be to determine the best time. I know that the first round of drone brood in my colonies (after winter oxalic dribble) contain almost zero mites, so removal at that time would only waste colony resources. This is also the generation of drones that mate with my raised queens. Drone brood removal at that time is of value to me as a way to eliminate drones from poorly-performing colonies from mating with my new queens–it allows to manage the genetics of the matings.

Later in the season, it may be appropriate to simply monitor the drone brood with a fork, during normal colony inspections (although this method as an indicator of mite population in the colony has been shown to be unreliable), and to cut out the drone brood when you see a significant mite infestation within.

The question for this strategy would be: Under what circumstances does the benefit of removing the mites in a frame of drone brood outweight the cost for the bees to replace it? If you have experience with this method, let me know your results.

Update 4/27/2011

This season we were not able to get our drone frames into our hives when we went to almonds.  Boy, did we notice a difference when we split the hives when we came home!  Since we started using drone trap frames, we had gotten used to seeing very little reworking of worker comb into drone comb on the brood frames.  Without the presence of the drone frames in the hives, the bees reworked large patches of worker comb into drone comb.  The effect, although not done as a controlled trial, was clearly noticable to all of us.  Could be a fluke, but I doubt it.

At this point in time, I really like having a drone trap frame in each hive year round, as it “manages” exactly where bees will build drone comb.  I then simply use a cappings fork to check the mite levels in the drone brood to see whether it is worth cutting it out.

Powdered sugar dusting

The second punch in our combination came from Finland. A few years ago, I came across an article in ABJ by the Finnish researcher, Dr. Kamran Fakhimzadeh (2000). He had been looking at various materials for dusting bees, with the intent to cause increased drop of the phoretic varroa mites. He hit upon powdered sugar as being both non contaminating to honey, and just the right size to clog up the mites’ feet. Pettis and Shimanuki (1999) had already proposed using dust in conjunction with a screened bottom board for varroa control. I mentioned the articles at our local bee meeting. One hobbyist member, Janet Brisson, was taken by the possibility of using sugar dusting and screened bottoms for mite control, and became a proselytizer for the method. I found the concept to be of interest, but the time-consuming method by which she and others went about applying the dust seemed totally unpractical to me.

Last year, at the club meeting, I asked for a show of hands as to what methods members were employing to control the mite. I was surprised that the majority were using sugar dusting! Finally, after seeing the method apparently being successful for two years, and reading Jerry Hayes (2004) promote the “Dowda” method in The Classroom, I felt that I needed to give sugar dusting a second look!

I did a few tests. I saw that sugar dusting sure did cause a lot of mites to drop, but it didn’t appear to affect subsequent stickyboard counts much, even after two weeks of repeated dustings! Something didn’t make sense. How could this method work if it didn’t affect sticky counts?

Then I started researching mite population dynamics, and realized the error of my thinking. Mites are only phoretic for about 5 days during broodrearing (a range of 4-15 days, dependent upon a number of factors (Harbo and Harris (2004)). Therefore, one would expect about a 20% turnover of phoretic mites every day, as older ones reenter brood cells, and new ones emerge. Knowing this, even if you had some new wonder chemical that killed 100% of the phoretic mites one day, you’d still have a 20% return of the phoretic mite population the next day, 40% by the second day, and back to the pretreatment level within a week! If you were to take a stickyboard count a week after the 100% kill, you would see zero effect from the dusting!

What I realized was that the problem wasn’t that sugar dusting didn’t work, but that I was not measuring its efficacy the right way! So I looked to the literature for measured levels of efficacy of an in-hive sugar dusting. To my surprise, there weren’t any. Fakhimzadeh had only measured the increase in daily mite drop and Aliano and Ellis (2005) had recorded a 75% mite drop only from caged bees. I contacted every researcher and beekeeper I could for an in-hive efficacy figure, but no one had one. So I collected the hard data myself, by dusting three test colonies (one, two, and three story), measuring the mite drop for the first hour, and then sacrificing all the bees in the colonies and washing the mites from them. I will write up a full version of the results when we complete testing, but in short, about a third of all phoretic mites in a colony drop in the first hour after dusting!

I now had a figure that I could use for crude modeling of the effect of repeated sugar dusting on mite population growth. I wanted to see if a mathematical model based upon the mite kill rate I measured would reflect the reports from the field. Since Fakhimzadeh and other beekeepers report that mites continue to fall at an increased rate for over 24 hours, I made the assumption that a good sugar dusting would kill 50% of the phoretic mites—a round number based upon a 33% initial kill, plus an arbitrarily assumed half again residual kill. These are working numbers subject to revision when we obtain more data.

Let me be clear at this point. I’m not about to recommend any varroa control treatment based upon mathematical modeling. What I’m curious to see, is whether the amount of mite drop caused by powdered sugar dusting could be mathematically expected to effect the mite control claimed by its proponents. So I called around to those who have been using the technique for over two years, and asked them for their records and observations. Some dusting practitioners had used ancillary treatments, such as drone brood trapping, or essential oils, so I allowed for those treatments. Their records indicated that: Initial dusting once a week for several weeks knocks mite levels way back, dusting twice a month keeps the mites at low levels, and dusting monthly (or even less frequently) keeps the mites at tolerable levels.

So, let’s see if crude mathematical modeling supports the field experience. I set up a simple mite population growth curve based upon a starting population of 100 mites, and a reasonable 2.4% daily mite growth rate (Martin 1998). Then I killed 1/6 of the total mite population at each dusting, based upon killing half of the one-third of the total mite population that is phoretic at any given time during the treatment period of March 1 through September 1. This model is very crude, and doesn’t account for amount of drone brood, multiple infestation, or other variables, and should only be used to give us a rough idea of the feasibility of the technique. I must admit, the results surprised me in how closely they reflected field experience! Clearly, powdered sugar dusting as a mite control measure has proven field efficacy, plus a mathematical model to support it.

Estimated effect of powdered sugar dusting

The estimated effect of powdered sugar dusting over a screened bottom on mite population growth, based upon a starting population of 100 mites, a daily growth rate of 2.4%, and an estimated kill of 50% of the phoretic mites per dusting treatment. Note that weekly dustings would result in a decrease in the mite population. These curves are based upon very crude math, and are only for general illustrative purposes, although they confirm field experience.

Note that the control curve reaches a devastating mite level by September 1st. Monthly dusting in this model keeps the mite population below a moderate threshold of 3000 mites, and bimonthly dusting keeps ‘em below 1000—a load that is considered acceptable by most all authorities. The weekly dustings actually decreased the mite population over the treatment period.

Not only that, but the illustrated curves likely underestimate the effect of sugar dusting, since even though it effectively kills only a sixth of the mites, the mites killed are those that would have been most likely to survive to reproduce. That is, once a mite is in the phoretic stage, its natural mortality rate is very low—about 0.6% (6 out of 1000) per day, as compared to the 20 –30% mortality of those first emerging from cells (Martin 1998). Although about two-thirds of the mites are under cappings and thereby protected from dusting treatments, that proportion is tempered by the fact that a quarter of them will not survive through emergence. This makes the mortality of the phoretic mites more important than their proportion might indicate. Recall from my discussion of mite population dynamics that that a female mite needs to average 2-3 reproductive cycles for varroa populations to grow at the pace that we see in the field. If sugar dusting knocks a mite down early in her life, she will be unable to complete multiple cycles. The surprising effectiveness of sugar dusting may due to its impact on the average number of reproductive cycles that a mite can complete.