Development of technologies for the control of varroa

Seventh Quarterly Report to National Beekeepers Association
Goodwin RM, Taylor MA, McBrydie HM and Cox HM October 2006

Executive SummaryMILESTONES FOR JULY 2006

1) Completed third crosses in breeding programme

EXTENSION

A summary of the results of this programme have been presented in thirteen seminars to various conferences, field days, workshops and meetings.

SELECTING HONEY BEES THAT ARE RESISTANT TO VARROA

Sixty-five genetic lines were established as colonies for the breeding programme and in autumn 2005 these were assessed for the level of suppression in mite numbers. The top ten lines that were selected to breed from displayed levels of mite suppression ranging from 17% to 56%.

Two hundred and sixty-seven single-drone inseminations were conducted on individual queen honey bees. Of these, 14 (5.2%) produced viable brood. This is less than expected from the first cross. The mite reproduction for the fourteen colonies was assessed and the levels of delayed suppressed mite reproduction (SMRD) were less than 20%. These queens were removed from the programme. The colonies were treated with Bayvarol®; to ensure they did not die from varroa over the winter. Fifty double-drone inseminations (semen from two drones) were conducted in April 2005 to ensure the lines were carried through to spring 2005. The survival of the second generation queens over winter was 34% which represented five of the ten lines. As there were insufficient mites to assess the queens SMRD in September 2005, two grafts from the 18 second-generation, double-drone inseminated colonies were conducted. Although frames containing drone cells had been put into each colony in August 2005, not all colonies had raised sufficient mature drones for inseminating. Some queens were artificially inseminated with two drones from either their second generation mothers or from a first generation breeder that had at least 17% SMRD.

A total of 146 inseminations were conducted. Only eight of these had suitable brood patterns to assess. The SMRD levels of the 18 over-wintered queens (generation 2) and eight spring 2005 inseminated queens (generation 3) were assessed. Eight of the queens representing five genetic lines displayed SMRD levels between 36-65%.

Thirteen new lines from two beekeepers were introduced into the programme in February 2006 and 10 from two additional beekeepers were added to the programme in autumn 2006. This was to increase the genetic pool of the closed mating population. The intention was to assess these lines for SMRD in spring 2006. However, as this project only has one more year of funding these lines will not be assessed and the resources will be used to increase the SMRD levels for the lines that are currently above 45%.

The intention at the end of this 2006/2007 year is to establish these lines in a natural system that may enable the lines to exist in a closed population.

IMPROVING THE EFFICACY OF FORMIC ACID TREATMENTS FOR VARROA CONTROL

Twenty-four, two-depth hives were treated with formic acid. Ten were treated with 250 ml of 65% formic acid, in a plastic pouch placed on the top bars. The other fourteen were treated with 2 L – 3.6 L of varying concentrations (0, 5, 10, 15, 20, 25, 30%) of formic acid. A plastic dish (320 x 320 mm) containing one of the seven concentrations was placed on the top bars of seven hives and on the floorboard of the other seven. The formic acid was topped up every four days when the sticky boards were replaced. After 20 days the pouches, dishes and sticky boards were removed. The base treatment was significantly more effective (4.1x) than the top treatment (P-value = 0.03). The effectiveness of the plastic pouches ranged from 5.6% to 54.1% varroa kill. This kill is equivalent to placing a 2% formic acid treatment in the base of a colony.

As the formic acid concentrations increase, there appears to be a negative effect on the bees and the brood. Whether this is a real effect is unknown.

A second trial to identify the relationships between evaporation rate, surface area and volume has been established. Different amounts (5 ml – 50 ml) of 20% formic acid were placed into six 50-ml containers and 20 ml was poured into a lid that had the same surface area as the containers. Each container was weighed at the start of the trial and will continue to be weighed throughout the trial.

A third trial commenced in November 2005 to establish the relationship between of formic acid evaporation and container head space. Thirty-two colonies were treated with one of eight concentrations between 10 and 25%. Each concentration was applied to one of the colonies in one of four treatment methods. These results will be analysed by July 2006.

The effect of head space above the formic acid on the evaporation of formic acid was assessed using five volumes (10, 20, 30, 40, 70 ml) of 85% formic acid. Twenty ml of formic acid was also placed into an upturned plastic lid (49 mm diameter, 13 mm high). The jars were then placed into a ventilated wooden box that was not temperature controlled. The weight of each jar was measured and recorded each day for four days and then again after 15 days.

The formic acid evaporated from all six containers, but the amount of evaporation varied depending on the distance from the top of the container. As the distance from the top of the container increased the amount of evaporation decreased. This is supported by 86% of 20 ml evaporating when placed in a lid with no head space compared with 30 % evaporation when placed in a container with a 40-ml head space.

A fourth trial using 34 hives commenced in November 2005. Two hives at the second site were selected as positive control hives and had Apistan® inserted at the recommended rate at the commencement of the trial. The remaining 32 colonies were randomly fitted with one of four treatment containers and filled with 2 L of formic acid (0, 10, 12.5, 15, 17.5, 20, 22.5, 25%):

1) Frame feeder

2) Top feeder

3) 50% of top feeder

4) Bucket

The formic acid when applied in the frame feeders at any concentration ≤25% did not kill more than 45% of the mites. The half top-feeders also did not result in effective varroa control as none of the concentrations killed more than 31% of the mites. The bucket feeder treatments killed up to 89% of the mites and the top feeders killed up to 83%. The kill rates observed from the bucket and top feeder application suggest that a large percentage of varroa in a hive can be killed when 25% concentration or less of formic acid is applied. However, the variation in percentage kill using these methods is too high.

A fifth trial commenced in March 2006 using 32 queen-right, two-super hives. Four randomly selected hives from each site were treated with Bayvarol® as controls. The other twelve hives at each site were treated with 2 L of 22.5% formic acid using a plastic top feeder for six weeks. A layer of nylon mesh (2 mm x 4 mm) was placed under the top feeder to prevent bees accessing the formic acid. A sheet of polythene was stapled over the top of the feeder to reduce the amount of vapours escaping under the lid. The hives treated with formic acid were randomly assigned a tin lid or a standard wood-lined lid.

There was no significant difference (P > 0.5) in formic acid evaporation between the hives with metal lids and those with wooden lids.

There was also no significant difference in percentage of varroa mortality between the hives treated with formic acid that had metal or wooden lids.

MPROVING THE EFFICACY OF THYMOL TREATMENTS FOR VARROA CONTROL

Thymol is an essential oil extracted from the thyme plant. Trials in New Zealand have shown that thymol crystals in dishes can be very effective at killing varroa. However, their effectiveness may vary considerably between hives in the same apiary. The other problem in using thymol is that a full treatment usually requires three visits to each hive.

The aim of this investigation was to improve the efficacy of thymol crystals against varroa.

Of the parameters measured, temperature and bee access to thymol appeared to be the most significant factors likely to explain the variability in treatment effectiveness. However, other parameters identified as affecting sublimation (surface area, edge effect, volume and the degree of compaction) that affect sublimation need to be taken into account when designing thymol-delivering apparatus.

Preventing honey bees from removing the thymol by limiting their access can probably be easily and cheaply achieved and is likely to produce significant improvements in the reliability and effectiveness of thymol treatments. Reducing the loss of the thymol would also reduce the cost of treatments and may also reduce the frequency with which dishes need to be refilled further reducing the cost.

Better positioning the dishes containing the thymol and possibly also utilising the brood temperatures may further increase the efficacy of thymol treatments.

Treating hives with thymol and oxalic acid at the same time did not significantly increase the efficacy of the thymol treatment. This lack of effect may have been because both act on phoretic mites.

A trial was established to determine the amount of thymol that should be used. Between 0.5 and 6.1 g of thymol sublimed over a four-week period. There was no relationship between the amount of thymol that evaporated and the percentage of varroa killed.

Restricting the access of bees to the thymol appeared to reduce the effectiveness to a point where the kill rates were not much higher than levels observed in untreated colonies. This suggests that bees may require access to the thymol for it to be effective. This would also explain why treating hives with both thymol and oxalic acid did not increase its effectiveness, as the honey bees in this case were also restricted from accessing the thymol. Whether the varroa mortality is a function of the passage of the thymol crystals through the brood nest or the presence of fallen thymol crystals on the floorboard is unknown.

Thus, although stopping honey bees removing the thymol appeared to be easily and cheaply solved and it looked like this was likely to produce significant improvements in the reliability and effectiveness of thymol treatments, the results of the oxalic acid trial cast doubt on the potential of restricting bee access to improve thymol treatment. Treating hives with oxalic acid at the same time as providing them with thymol screened to prevent bee access did not improve the efficacy of the thymol. The kill rate of the thymol was also much lower than that seen in other trials. Restricting the access of bees to the thymol therefore appeared to reduce the effectiveness of the thymol. Whether the varroa mortality was a function of the passage of the thymol crystals through the brood nest or the presence of fallen thymol crystals on the floorboard was unknown. If the presence of thymol crystals on the floorboard was important then the removal of sticky boards might be been reducing the effectiveness of the treatments.

When a trial was conducted to test this without using sticky boards, the results were less clear. There was no difference in the mortality irrespective of whether bees had access to the thymol crystals or not. Crushing the crystals did, however, appear to improve the effectiveness of the treatment. There was a relationship between the amount of thymol removed from the dishes and the adjusted estimate of the number of mites killed.

This suggest that increasing the amount of thymol removed from the dishes (through bee removal or sublimation) should improve the varroa kill as long as the amount of thymol does not get high enough to interfere with the bees. The results of putting large amounts of thymol into hives did not, however, support this. This needs to be investigated further.

RATE RESPONSE CURVE FOR APISTAN® AND BAYVAROL®

Apistan®

The trial to establish a rate response curve for Apistan® has been completed. The results indicate that the recommended rates used are far higher than the rates required to control varroa. A high level of control was still observed when a rate as low as 3.1 percent of the recommended rate was assessed. This indicates that the practice of using Apistan® at 50% of the recommended rate may be carried out without risking the acceleration of resistance development through increased survival of tolerant mites. The effect of reducing the strip distribution has not been assessed. As resistance of varroa to fluvalinate develops further over time this situation may change.

Bayvarol®

The trial outlined above was repeated on 7 October 2005 using Bayvarol®. Mortality of varroa when exposed to the recommended rate of Bayvarol® was slightly higher than that observed for the recommended rate of Apistan®. The mean percentage of varroa killed remained above 90%, even when rates were reduced to 30% of the recommended rate. However, there was a sharp decrease in varroa mortality once the application rate decreased below ~20% of the recommended concentration.

The effectiveness of Bayvarol® was reduced to a greater degree than Apistan® as the rates decreased below 20% of the recommended concentration. This is not unexpected as Apistan® has 824 mg of fluvalinate per strip (1648 mg/brood box) compared with Bayvarol with 3.6 mg of flumethrin per strip (7.2 mg/brood box). The observed decrease in effectiveness suggests there is a greater risk of varroa developing resistance to Bayvarol® if the amount of Bayvarol® was decreased than for Apistan®.

INVESTIGATING THE POTENTIAL FOR USING METARHIZIUM ANISOPLIAE AS A BIOLOGICAL CONTROL AGENT AGAINST VARROA

Three Metarhizium isolates were sourced from the Landcare Research collection and six isolates from the AgResearch collection. Three isolates were tested on caged bees and on varroa. The isolates had no effect on honey bee survival, but they killed almost 100% of varroa within 24 h.

Eighteen colonies were divided between three treatments. The treatments were two treatments of 5 g of Metarhizium sprinkled across the top bars, eight days apart, two treatments of 5 g of dead Metarhizium sprinkled across the top bars, eight days apart and no treatment.

Significantly (P < 0.03) more dead mites were deposited on the sticky boards from the hives treated with live Metarhizium than the untreated hives. Although more mites were present on sticky boards in the hives treated with dead Metarhizium than in the untreated colonies the difference was not significant (P = 0.11).

The live Metarhizium treatment also significantly (P = 0.017) increased the percentage of the varroa in the hives that died during the treatment period. The untreated and dead Metarhizium treatments were not significantly different (P = 0.25).

The first steps in a laboratory bioassay for determining the effectiveness of Metarhizium strains have been completed. Further work is required to improve the survival of varroa used as controls.

Completed length of spring treatment trial

The aim of this trial was to determine whether the treatment time over spring can be reduced compared with those used in the autumn period. There was no significant difference (P = 0.066) between treatments at the 17 January 2005 assessment date. However because of the lateness of the season it was decided not to finish the trial at this time but to continue it for a further two months. Hives which were treated for the shortest time period (two weeks) were removed from the trial as several had more than 20 varroa per sample.

There was no significant difference (P < 0.97) in the number of mites per bee or the numbers of mites in the hives when the trial was discontinued on 2 March 2005.

This suggests that four-week treatments are as effective as eight-week treatments. This has the advantage that there is less time for varroa to develop resistance, and less time for residues to develop and only half the product cost. The reduced product cost assumes that strips can be stored and reused for the following season. Whereas this may be possible for Apistan® and Bayvarol® because of the long-term stability of fluvalinate and flumethrin, it may not be possible for Apivar® because armitraz is less stable.

Completed treating colonies for varroa while re-queening with cells

The first experiment was designed to evaluate the effect of treating varroa on queen survival. Treatment of hives with oxalic acid after re-queening had no effect on subsequent survival or mating success of queens.

To determine the length of the broodless period following re-queening, twenty beekeepers agreed to re-queen 10 hives each and monitor the length of any broodless period. Despite reminders, only 10 of these beekeepers completed this trial. This was in part due to the very poor spring weather conditions. Of the 93 hives that were re-queened, 62.3% had no capped brood when checked after 25 days.

The third experiment was designed to evaluate the efficacy of varroa treatments applied to broodless colonies. The application of oxalic acid during the broodless period on average killed 61.2% (S.E. = 19.8.) of varroa while the Bayvarol® treatment killed 89.3% (S.E. = 8.26). The relatively low kill rate using oxalic acid was unexpected, as the accepted reason for a low kill rate by oxalic acid is that varroa that are sealed in brood cells are not affected. If this trial was to be repeated, larger volumes of oxalic acid should be used.

For further information, please contact:
Mark Goodwin
The Horticulture and Food Research Institute of New Zealand Ltd
HortResearch Ruakura
Private Bag 3123
Waikato Mail Centre
Hamilton 3240
NEW ZEALAND
Tel: +64-7-858 4650
Fax: +64-7-858 4700
Email: mgoodwin@hortresearch.co.nz