Evaluating a Heat Therapeutic Control of the Honey Bee Mite Varroa Destructor

Description of the heat-treatment apparatus and bee handling procedure
The apparatus used in this project was, as proposed, built similarly to a design of a home-made Russian apparatus documented by the project coordinator in use in Uzbekistan in 1994. It consists of a portable wooden cabinet which accommodates a standard 1500 watt "milk house" type heater that stands on the floor of the cabinet and a wire mesh cylinder or "cage" for suspending the bees in the upper portion of the cabinet. Plexiglass windows were fitted into the top and sides of the cabinet for ventilation and observation. Bees are shaken, brushed and/or blown from the hives frames into the heat-treatment cage via a large funnel and thus transferred into the cabinet for heating. The actual heating can take less than 15 minutes per hive while the handling of the bees can take twice as long.

The first trials with the heat-treatment apparatus, in August 1996, preliminary to the first efficiency study, showed that an accurate and practical indicator of effective temperatures is the falling of mites from the caged bees. By observing the thermometer and mites falling onto a paper lined cookie sheet, the operator was easily able to obtain effective mite-fall temperatures within the cabinet. Then by turning the heater off or on and opening or closing either of two sliding Plexiglass windows he was able to maintain mite-fall temperatures or begin cooling All of the treatments performed during the course of this project followed this same basic procedure.

Objective One: Efficiency Study
Efficiency study heat-treatments were performed, as were all subsequent treatments, in the following fashion. The bees were loaded into the treatment cage, the cage inserted into the cabinet which had been pre-heated to between 25-30˚C and the electrical circuit to the heater closed. When an accurate full-immersion type glass thermometer positioned in front of and near the top exterior of the cage showed 47 degrees Celsius (116.6˚ F), Varroa mites began to fall from the caged bees. Other thermometers and thermocouples positioned inside and below the cage sometimes showed a differential of 10 degrees or more, but as long as the reading at the primary thermometer stayed above 47-48 degrees, depending slightly on ambient temperature, mites fell at a more or less steady rate for 2-6 minutes, never as long as 10 minutes. A cookie sheet lined with white paper was positioned on a screen shelf below the cage where it was easily visible through a Plexiglass window. When mite fall was observed to cease and a little agitation of the cage failed to produce more mites, the heat was turned off and ventilation provided for cooling.

For the purposes of this efficiency study 4 colonies ranging in size from 456 to 1133 grams (1 - 2.5 pounds) were heat-treated and the bees returned to their hives over a "sticky board" so that additional mite fall and bee mortality could be measured. The hives to which the bees were returned where brood free and the comb reasonably clear of mites. Egress or ingress of bees was prevented by covering the hive entrances with wire mesh. Two plastic strips coated with 10% fluvalinate (commercial name: Apistan) were inserted and mite fall counted over the next 9 - 14 days. This control treatment of known very high efficiency, administered to single-chamber brood-free hives was expected to remove virtually all the remaining varroa mites. Efficiency of the heat-treatments was calculated by dividing the number of mites detached during heat-treatment by total mites and ranged from 81.9 to 94.6.%

The efficiency of mite removal with two additional colonies was similarly measured except that instead of a control treatment with Apistan the bees were sacrificed and all remaining varroa mites counted using a 30 minute agitated alcohol bath and straining (DeJong et al '82). The efficiencies of these two treatments were 90.6 and 97.9% for colonies weighing 1420 and 1498 grams (3.13 and 3.30 Lbs).

Contrary to expectations that treatment temperatures and durations would need to be minutely adjusted to maintain efficiency in treatment of different sample sizes, the 47-50 degree temperature range worked for all sample sizes below 1588 grams (3.5 pounds) as long as the cage was not over filled and other operating procedures observed.

During heat-treatment trials, preliminary to the efficiency study, it became apparent that certain factors grossly affected the efficiency of treatments. For example if the mixed adult bee population of a large varroa infested colony was treated in successive batches it became obvious that while a cage filled with bees so as to allow ample space for air circulation produced an impressive mite fall, a cage overfilled with bees was slow to heat, detached very few mites and was more likely to result in and damage or death to bees. If the treatment cage was oversized for the cabinet, mite fall was also grossly reduced. For the purposes of the efficiency study and subsequent treatments performed as part of this project an 11 inch diameter X 16 inch long cage was used.

If over-heating occurs, with the primary thermometer readings above 49-50 degrees, some bees began to regurgitate a clear liquid, became excessively lethargic, and more likely to clump together in collective helplessness made worse by the cohesive effects of the regurgitated liquid. If the cage was overfilled or the bees in the cage were in other ways allowed to clump together during heating far fewer mites fell than with the treatment of loosely distributed bees from the same colony. So whether clumping is allowed at the beginning of critical heating or results from over heating, in both cases bees are undesirably stressed and the efficiency noticeably reduced.

A more detailed description of bee "clumping", cage size, and capacity follows under findings for "Objective two: Improvements".

Objective Two: Improvements to the Heat-treatment Apparatus and Procedure
Experience with heat-treating bees enabled the project participants to effect improvements to the heat-treatment apparatus. It was discovered that bee handling and heat-treating was greatly facilitated through preparatory manipulation/inspections and choice of when to treat.

(a) Cages and Cage Size.       Three cages of cylindrical mesh (mesh size: 8 x 8 openings per inch) were built to the following dimensions: 8 inches diameter x 16 inches long, 11 x 16 and 14 x 19. Samples ranged from 170 grams (6 oz) up to 3.63 kg (8 pounds) of bees per cage. If the cage was filled beyond its capacity excessive clumping occurred as the surface area of the inside of the cage was insufficient to allow a grasp for a large number of bees who ended up falling to the bottom of the cage in a clump. The greater the number of bees in a treatment the greater the probability of treatment and post treatment clumping.

Guidelines for determining appropriate cage sizes should take into account bee distribution on the mesh; the surface area of the cage should be equal to or a little more than that which the bees can cover evenly. A few trials with samples not exceeding 1360 grams (3 pounds) in the 11 inch cage proved effective, but above 1587.6 grams (3.5 pounds) clumping was more difficult to prevent. So maximum number of caged bees per square inch is as follows: 1360 gr bees ÷ .121 grams per bee (average spring bee weights) = 11240 bees, 11240 bees ÷ 743 square inches per 11 inch x 16 inch cage = 15 bees per square inch. This 15 bees per square inch density was greater than expected; subsequent observation suggested the high number of bees which can manage to hold on to a square inch of wire mesh actually involves the remarkable ability of the bees to interlock with each other and the cage while continuing to move about the surface.

In early spring the average weight of the project colonies was approx. 680 gr.(1.5 lbs). It was found that with a second 11 inch cage it was possible to fill one cage from one colony while still treating the first one or to treat as large as a 2720 gr.(6 lb) colony by filling each cage with 1360 gr (3 lbs) of bees and heating each cage in turn.

(b) Cage construction materials.       A cage frame built entirely of metal will work very well but requires welding. The 11 x 16 inch cages built in the course of this project did not require access to welding skills or equipment but instead were built using 1/4 inch threaded rod and 1/2 inch plywood. Experience with different meshes showed that a common type of galvanized 8 x 8 mesh seemed more thermo-transparent than a heavier gage 8 x 8 steel mesh.

(c) Minimizing clumping of bees during treatment.       Since in most cases--assuming the cages are not overfilled with bees--clumping can be easily prevented by manipulating the cage with its axis rods which protrude from the cabinet, it was not attempted to quantify the possible reduced efficiency caused by clumping. However two causes for clumping induced inefficiency can be supposed and are included as follows: 

Firstly, excessive clumping of bees in the cage seems to prevent even heating from both convective and radiant sources. Some bees seem overheated while others are shielded and under heated. Secondly, it seems reasonable to suppose that if mites are heated and dislodged despite clumping together of bees in the treatment cage, fewer detached mites can fall free of the clump especially if the clump is sticky. The mites may survive to reattach to another host as the bees and mites are cooled. This is supported by the observation during all of the 24 heat-treatments performed during this project that some mites falling from the treatment cage during heating remained obviously alive and animated. Even after contact with the catchment pan, which reaches temperatures above 50 degrees, some mites survived the treatment if removed promptly and would readily re-attach to any prospective host bee introduced into their container.

(d) Avoiding post treatment clumping, cooling techniques.       By trial and error it became apparent that clumping after the treatment presented as much of a threat to treatment success as clumping of bees during treatment. Upon reintroducing a colony to the hive it was a challenge to distribute helplessly lethargic bees without damaging many of them especially if they were sticky. An improperly re-introduced cluster can bake itself especially if the hive is closed up before the cluster has a chance to eliminate excess heat. This happened with a preliminary treatment colony which, during a postmortem 24 hours later, gave off a smell of baked bees and was still unusually warm to the touch. The first step to avoiding this problem while the bees were still in the cabinet was to stop heating the bees immediately after the mite fall ceased. Next it was found that removing one or two frames and spacing the remaining frames made it easier to coax the reintroduced bees down in between the frames where clustering was limited by the combs.

These techniques helped reduce or even eliminate post treatment bee mortality and were employed during every treatment performed during the principle research study. Consistent with the Russian-Uzbek method the caged, treated bees were usually allowed to eliminate excess heat while still inside the cabinet with the vents open and the heater off. Later it was also found that in many cases, depending on ambient temperatures, additional pre-cooling of caged bees outside the cabinet was helpful.

(e) Transferring Bees from the Hive to the Treatment Cage.       It is the handling of bees from the hive into the treatment cage that of all the steps takes the most time and effort. Two types of funnel were employed for this purpose. The first funnel was only large enough to accommodate in its mouth a frame covered with bees and conduct them down into the treatment cage. The bees were dislodged by shaking, brushing, and sometimes blowing with the exhaust of a shop vacuum. Handling techniques of this sort are routinely employed by beekeepers when harvesting honey supers and making-up packages of bees, and can be done with little damage to bees. In cold weather a blower should not be used as it can excessively chill the bees.

The single frame funnel design used in this project had to be modified from the Russian-Uzbek model whose dimensions accommodate a single frame 30 % larger than the American or Langstroth type frame. The multi storied Langstroth type of hive used in the US with its ten frames per story or "chamber" differs from the Russian type in both size of frame and size and configuration of chamber. The Russian hive is a single chamber which is used with 18 or more large frames These differences between American and Russian hives not only require a 30% smaller single frame funnel but additional modifications.

Because of differing hive configurations the winter cluster of bees might occupy only a few frames in the Russian hive but portions of as many as 12 frames in the Langstroth hive. This means that the heat therapist of a Langstroth hive must find a way to handle all those extra frames. In this project a second funnel type was built large enough to support an entire brood chamber. A screen was fitted over the top of the chamber and a blower employed to move the bees off the frames down the funnel into the treatment cage without the necessity of handling frames individually.

(f) Preparation of Colonies for Heat-treatments.       Consolidate Bees onto a Minimum of Frames in Advance of Treatment. Another way to facilitate handling of frames of bees is to prevent them from occupying more than a maximum of 8-9 frames. Consolidation of bees onto selected frames for the purpose of heat-treatment greatly facilitated heat-treatments conducted during the course of this project. A typical triangle type bee escape board was inserted between the upper and lower chambers of the hive a day or two in advance of the scheduled heat-treatment. A great advantage to advance consolidation of the cluster in this way is that the beekeeper can select mostly empty frames for the consolidation area thus easing the shaking and brushing process which can be complicated by the weight of honey comb and the spilling of loose nectar or honey onto the bees prior to treatment. Care was taken to make sure at least one frame of honey and pollen was available to the consolidated cluster for their consumption during the 24-48 hours leading up to the treatment.

About Queens       It is a practice among some of the Central Asia practitioners of Heat therapy, and it seems a sensible one, to protect the queen from potential damage due to handling or heat by removing her to a pocket sized holding cage for safe keeping until the heat-treatment is finished. In this project the queen was caged within the hive at the same time as colony was consolidated. At the time of treatment she was kept in a warm pocket.

Preparation of colonies presented an opportunity to combine several seasonal management tasks into one or two detailed inspections. A frame to frame search for the queen familiarized the beekeeper with the quantity and quality of honey and pollen stores necessary for wintering or spring survival. Disease detection can also be performed. As the colony is undergoing treatment its hive can be cleaned of dead bees and debris, the chambers configured with selected frames taking into account the colony's winter need for sufficient honey, pollen stores and clustering space.

(g) Choosing When to Treat.       Broodless Periods Common in Fall and Winter. Varroa mites reproduce in the brood cells on pupating bees where they are protected from most mite control methods. Therefore it is desirable to treat colonies for varroa during those few, geographically variable, periods when brood is not present in the hives. A Univ. of Guelph Study showed November broodless periods among 28 wintered colonies(Szabo '93). During the heat-treatment study we found significant cold weather periods of near or total broodlessness (in fall, winter and spring of 1996-97) among the 32 colonies in this study.

Cold Weather Handling. Some experimental chemotherapies for varroatosis lose effectiveness during the cold weather when broodless periods would otherwise make treatment desirable. By contrast, this study found multiple advantages to heat-treating colonies during cold broodless periods. Most beekeepers do not need to disturb colonies during cold weather and prudently avoid doing so under the assumption that colony health will suffer. During the course of this study it was occasionally necessary to inspect colonies and handle bees in air temperatures below freezing. In the absence of brood, the colonies showed no obvious long term effects of exposure to freezing air temperatures for the roughly 5-30 minutes it took to inspect a colony or prepare it for heat-treatment: with care, hives were opened and the bees emptied into cages for successful treatment without causing significant bee mortality.

If sufficiently disturbed while being handled for heat-treatment many bees would fly and escape heating thus reducing the efficiency of heat therapy. It was preferable to prevent bees from flying during the pre-treatment handling. Cold weather handling was an advantage since the bees in this study would fly only reluctantly in temperatures much below 50-55˚F (10-13˚C). However, wind, rough handling or the accumulation of alarm pheromone on the beekeeper and equipment could induce bees to fly regardless of low air temperatures resulting in undesirably high bee mortality.

Nocturnal Treatments. The fact that bees do not fly after dark provided a solution to the problem of some bees flying before they could be placed in a treatment cage. All treatments recorded during this project were conducted at dusk or in the dark. The field researchers wore head-lamps and a low wattage lamp was attached to the treatment cabinet.

(h) Periodic Cleaning of Apparatus.       The cages should be cleaned with a brush, soap and hot water between sessions or apiaries in order to keep the cage free of debris that might decrease thermo-transparency, remove accumulated alarm pheromones and other smells that may unduly disturb the bees, and limit the potential transmission of disease from apiary to apiary. Heating of the empty cage and box above 50˚C for 24 hours or more will kill Nosema spores (Sammataro and Avitable 1986) and possibly other disease organisms. Such procedures were practiced during the course of this project and are recommended for all bee thermo-therapists.

Objective Three: Winter survival after fall treatments
It was originally planned to obtain the livestock for this project already infested with Varroa mites and in the interest of moderating costs as much as possible perform the first set of treatments in September 1996 so as to generate winter survival data early in 1997. Contrary to the bee supplier's original stated intentions he treated all his stock for varroa using Apistan. It was too late to find other sources of untreated colonies so the fall treatment was cancelled; the senior investigator and entomological advisor modified the project design so that 1/2 instead of 1/4 of the 32 colonies would be treated in spring. The result was a simplified experiment with increased repetitions.

After acquisition in September 1996, all 32 colonies were collected into a holding yard where they were weighed, fed sugar syrup. The colonies were seeded with brood and bees from some of the coordinator's colonies which were heavily infested with varroa. Special care was taken to seed equally and randomly assign colonies to sites (see table 5).

Objective Four. Summer, fall assessments of treated and control colonies
With the forced cancellation of the fall 1996 treatments the objective was to compare progress of spring 1997 treated colonies and control colonies over the summer of 1997 through November. Both groups originally included 16 colonies; the treated group lost 1 robust colony to late winter starvation and the control group lost two colonies which exhibited mite associated symptoms over winter-spring. Such losses are within the normal 10% experienced by healthy colonies.

Between 28 March and 5 April 1997 the fourteen surviving colonies randomly assigned to the treatment group were heat-treated in situ. All colonies were managed in conventional fashion over the following spring and summer except that detailed inspections quantified cluster size, brood, honey, pollen on a regular basis.

The first outcome measure to distinguish treated from control colonies were the sticky board counts begun in July and August. The sticky boards consisted of oil coated paper, separated from the bees by 1/8 inch mesh, placed on the floors of the colonies to show the number of varroa mites naturally falling in each colony (see table 3). These counts were conducted over four 10 day to two week periods July - November 1997.As is typical with varroa infestations the mite populations began to boom in August. Analysis of sticky board data showed that the control group produced 7.2 times more mean mites per day than the treated group.

In second or third year infestations, untreated colonies of honeybees often crash dramatically and die after the annual varroa populations boom. This study was not funded for the two-three years necessary to compare control and heat-treated colonies over the full two to three year course of varroa disease. Advanced stages of varroatosis include an aggregate of little understood symptoms associated with viruses which is called parasitic mite syndrome (PMS). These symptoms appeared in several colonies but did not occur in one group more than the other.

Honey production by the control colonies was higher than the treated colonies, but beekeeper experience with varroa infestations suggests that honey production is often an unreliable indicator of varroa infestation , and in this study honey production was measured as an indicator of general colony vitality and forage opportunities. Other factors probably affecting honey production were bear damage to three of the treated colonies and the very poor forage opportunities at a treated site that had been expected to provide good forage (see table 2).

Objective Five. Tracheal mites (Acarapis woodi)
In March 1997 a total of fifty bees from each site were collected from the inner covers of the colonies and examined for tracheal mites to establish base line data prior to heat-treatments. The bees were dissected and examined under 40x dissecting scopes according to established procedure for tracheal mite detection. On average each yard showed one bee in 50 had tracheal mites at a time of the year when tracheal mite infestations are expected to be concentrated. With such low counts a specialist indicated we should perform these examinations every month. The projected labor allotment for this new work was so great as to be impossible within the financial scope of this project, and no further investigation of the effects of heat therapy on tracheal mites was attempted.