Tower-hive configuration for the maximization of honey yields and increased efficacy of easy drone brood removal

Final Report for ONE07-076

Project Type: Partnership
Funds awarded in 2007: $9,461.00
Projected End Date: 12/31/2008
Region: Northeast
State: Pennsylvania
Project Leader:
Dennis vanEngelsdorp
University Maryland
Expand All

Project Information


Capped drone brood removal is a varroa mite control technique that reduces the need for chemical treatments. This varroa mite control technique has not been readily accepted by many beekeepers because heavy honey boxes must be removed to access part of the hive where drone brood is produced. This study investigated a way of managing hives that makes it easier to access the brood nest and the contained drone brood, making drone brood removal a more viable option.

In this study we demonstrated that the management of honey bee colonies in a two-queen tower configuration and utilizing drone brood removal reduced varroa mite population growth over the course of the summer. Colonies managed in a tower configuration contained fewer mites in early September than those that were managed normally. These colonies were also able to produce equivalent amounts of honey as those managed normally. While this system would not likely permit beekeepers to forgo alternative mite treatment methods, it would keep mite levels below economic injury levels for a longer period of time.


The most destructive pest of honey bees, Apis mellifera, is the varroa mite, Varroa destructor. These external parasitic mites feed on the blood of adult and immature bees and reproduce within cells with developing bees. Mite infestations lead to increased transmission of viruses, deformed and underweight bees, reduced honey production and, if not controlled, colony death. The loss of great numbers of colonies has resulted from varroa mite infestation.

Control measures are necessary to prevent the death of honey bee colonies. Chemical treatments can leave residues in bee products, such as honey and wax (Bogdanov et al. 1998), and have led to mite resistance (Elzen et al. 1998, Pettis 2004, Lodesani and Costa 2005). A more sustainable option is the use of integrated pest management (IPM), which involves using a combination of several mite reduction techniques which can include chemical and cultural controls. Drone brood removal is a cultural control technique that may be useful as part of an IPM program. Because varroa mites prefer to reproduce in drone cells (where male bees are developing) rather than worker cells (where female bees are developing), drone brood can act as a sink for varroa mites. Removing drone brood infested with varroa mites can be a very effective control measure (Lavagnino and Marletto 1995; Calderone 2005). To remove frames of drone brood from traditionally-managed colonies, however, heavy boxes full of honey must be removed from colonies to access the brood nest. If there were a way of managing hives so it is easier to access the brood nest, drone brood removal would be a much more feasible option.

Natural honey bee colonies have only one queen, but some beekeepers use a two-queen system (Cale 1963; Hogg 1983) where colonies are maintained with two laying queens in one hive. Using this system, the two queens are kept in different brood boxes with a screen between them that allows worker bees to pass through but not queens, which are larger. This management technique is thought to increase the honey bee population and, subsequently, the honey yield (Winston and Mitchell 1986). In addition, if one queen dies, the colony survives without an interruption in the addition of young bees.
In this study, we used a new hive configuration that took advantage of a two-queen system and permitted easy drone brood removal.

Project Objectives:

Our goal was to determine whether this system would allow beekeepers to keep the varroa mite population from dramatically increasing over the summer without using chemicals. We also compared honey production of colonies kept in this configuration to that of colonies kept in a “normal” configuration.


Click linked name(s) to expand/collapse or show everyone's info
  • Craig Cella


Materials and methods:

Four beekeepers conducted this experiment over 2 years (2007-2008) using 76 honey bee colonies that were randomly assigned to one of two treatments: 1) normal management with two brood boxes, a queen excluder, and honey boxes (n = 36) colonies) or 2) management in a “tower” arrangement (n = 40 colonies). The tower arrangement consisted of two colonies with two brood boxes each placed side-by-side with a queen excluder (mesh allowing worker bees, but not queen bees, to pass through) and honey boxes over the center of the two colonies (Fig. 1). The exposed half of top of the upper brood box was covered by a half-size lid. This setup allowed easy access to half of the frames in the upper brood box of each colony (Fig. 2), creating a two-queen system and enabling us to add and remove drone comb for mite trapping without having to remove honey boxes. In addition, part of the brood nest could easily be inspected.

Drone comb (a frame containing cells in which male bees develop) was added to the top box of each colony in the tower configuration in June (green frame in Fig. 2). Periodically over the summer, this drone comb was removed and replaced with new comb. The comb that was removed was frozen to kill the varroa mites in cells. Timing of this removal varied as each beekeeper conducted the experiment. Empty boxes were placed over a queen excluder as needed for all colonies, whether managed normally or in a tower configuration.

When possible, brood of the proper age from the removed frame was examined for the presence of varroa mites. The goal was to examine 50 capped drone brood cells per frame for the presence of varroa mites to quantify how many mites were being removed. The total amount of drone brood present on the frame was also estimated.
The varroa mite population was monitored by using sticky boards. Sticky boards are large pieces of thick white paper covered with petroleum jelly that are placed on the floor of each hive. These devices collect mites that fall from the colony and measure what we call natural mite drop. The number of mites that fall onto the sticky board each day gives a good indication of the size of the mite population in the entire colony. This was measured in all colonies at the beginning of the season (mid-June) and at the end of the season (mid-September) to determine how the mite population changed. In addition, 2 beekeepers monitored the mite population throughout the entire season. For data analysis, each hive was considered separately, whether paired in a tower configuration or whether in the control group.
In the fall, the honey was harvested from the colonies and the removed boxes were weighed to determine how much honey each colony produced. In addition, one beekeeper recorded his observations of the presence of the queen in his colonies at that time.

Statistical analysis: Mite drop data from the sticky board samples taken at beginning and end of the experiment were analyzed using a split-plot analysis of variance design (PROC GLM; SAS Institute Inc. 1999). Treatment was the main plot factor and time was the subplot factor. Fisher’s exact test was used to determine whether treatment affected queen or colony survival using contingency tables set up with three columns (treatment, status, and count).

Research results and discussion:

In the normal colonies (without having drone comb removed), average daily varroa mite drop increased from 1.23 ± 0.23 mites per day (mean ± s.e.) at the beginning of the study (mid-June) to 14.69 ± 5.53 mites per day at the end of the study (mid-September). In contrast, mite drop in the tower colonies, which had drone comb removed, increased from 1.08 ± 0.26 to 3.78 ± 0.70 mites per day. Thus, the change in mite drop over time was affected by treatment (F = 4.11; df = 1, 148; P = 0.0444). The course of the increase in mite drop over the duration of the experiment is presented in figure 3.

In addition, the colonies from the two treatments produced equivalent amounts of honey with the normal colonies producing an average of 19.9 ± 4.3 kg (n = 12 colonies) and the tower colonies producing 17.7 ± 2.2 kg (n = 22) each or 35.4 ± 4.4 kg per pair (F = 0.26; df = 1, 32; P = 0.6119). On average, each drone frame removed from a tower colony contained 1,650 capped drone cells. Between 0 and 9 % of the drone cells that were opened contained varroa mites.

One beekeeper recorded his observations of the presence of the queen in his colonies at the end of the experimental period. In all, 2 of 8 colonies in the tower treatment and 1 of 8 colonies in the control died before the end of the experiment. In addition, 1 queen was lost in the tower treatment. No differences between treatments were found (P > 0.05).

Management of honey bee colonies in a tower configuration utilizing drone brood removal and a two-queen system reduced varroa mite population growth over the course of the summer. Colonies managed in a tower configuration contained fewer mites in September than those that were managed normally. While this system would not likely allow beekeepers to completely eliminate the need for alternative mite treatment methods, it would keep mite levels below economic injury levels (the point at which harm to the colony is more expensive than the cost of a mite control treatment) for a longer period of time.

The tower colonies went into the fall with much lower varroa mite populations than the control colonies. These findings are similar to those reported by Calderone (2005) who removed drone brood from colonies managed in a normal configuration. In this study, mite populations increased throughout the summer, as expected. This increase was significantly reduced when drone brood removal was used. If mite population growth is slowed, beekeepers can confidently delay treatment until the end of the honey flow without compromising the health of the bees.

This management technique was easy to use and allowed bees to collect an equivalent amount of honey as compared to normally-managed colonies. The colonies that were monitored for honey production had been created using honey bee “packages”. Packages are colonies without hives that are shipped to the beekeeper that then puts them in a hive. The packages in this study were put into hives with frames that needed to be manipulated by the bees before honey or brood could be produced. This process is energetically expensive and, generally, leads to reduced honey production (Hepburn et al. 1984). If drawn comb were used instead of foundation, the bees would not have had to expend energy on wax production. Furthermore, previous work showed that the removal of drone comb increases honey production, likely due to the elimination of the need to care for costly drone adults (Seeley 2002), so further work in this area is necessary to determine whether colonies managed using drone brood removal may in fact be capable of producing more honey than colonies managed without using this technique.
The two-queen system did not increase or decrease the honey yield. In contrast, Gutierrez and Rebolledo (2000) were able to increase honey production using a two-queen system in Chile. In addition, Winston and Mitchell (1986) found that colonies managed with two queens contained more brood and bees than those managed with one queen, but found that the labor involved in the system negated these positive outcomes. Using the tower configuration, labor was greatly decreased, allowing the two-queen system to be profitable.

Research conclusions:

We feel that the addition of drone brood removal to an IPM program using a tower configuration will be advantageous. The use of chemical acaricides will be reduced and can be delayed due to decreased mite reproduction. The tower configuration allowed for easy removal of drone brood, allowing this cultural control method to be used without added labor. We recommend the use of this system by beekeepers as part of an integrated varroa mite control program.

Participation Summary

Education & Outreach Activities and Participation Summary

Participation Summary:

Education/outreach description:

The results of this study are being published in Volume 1 of the new journal The Science of Bee Culture. In addition, the results have been shared at both PA state and national and international bee research meetings.

Project Outcomes

Project outcomes:

Reducing beekeepers’ reliance on chemical controls for varroa mites will save time and labor costs. In addition, the tower configuration allows beekeepers to easily remove drone brood at any time of year. This is a great advantage over all other mite control techniques which require that they be implemented when honey is not being produced.

Farmer Adoption

We do not have any official information about how many beekeepers are using this technique, but have been told anecdotally that some beekeepers are trying it.

Any opinions, findings, conclusions, or recommendations expressed in this publication are those of the author(s) and do not necessarily reflect the view of the U.S. Department of Agriculture or SARE.