EU_report

The Commission Of the European Communities, Director General for Agriculture (DG VI.F.II.3) has given support 1998-1999 for a Concerted Action 3686, entitled "Coordination in Europe of research on integrated control of Varroa mites in honey bee colonies"

Varroa mites in European honey bee colonies must be controlled if colonies are to survive. The most common treatment regimes involve use of synthetic acaricides inside the bee colonies. Although effective, these control methods have created new problems for the beekeeping in Europe, such as acaricide residues in bee products and, lately, acaricide resistant mites.

Many studies demonstrate that several organic acids, already present in small amounts in bee hives (lactic, formic, oxalic acid) can be used to control varroa mites. There is also growing evidence that some etheric oils and some common food additives, as well as biotechnical management, can be used effectively for mite control. However, these ecological alternatives to synthetic acaricides are generally more laborious to use and show variation in efficacy depending on external factors and colony condition. Thus, they need more attention and knowledge on behalf of the beekeeper to be sufficiently effective. From practice it is also evident that methods that work well in one region may be inappropriate in another with a different climate.

Thus, there is a large need for a network of scientists in this field, capable of identifying the research needs, to harmonize efficacy evaluation protocols, and disseminating research results and recommendations in a comprehensive form that may be used as guidelines for beekeepers in Europe.

The main objective of this Concerted Action (CA) is to coordinate research efforts in Europe on integrated varroa control and to disseminate information to the beekeeping practice on how pesticide use in beekeeping is best avoided.

* Establish a European network of scientists for efficient exchange of information on methods that minimize the use of pesticides for control of the parasitic mite Varroa jacobsoni in honey bee colonies.

Texts in this compendium are working material presented and discussed by invited European scientists, advisors and legislators at the CA meeting at the Agricultural Research Centre-Ghent, Belgium, November 13-14, 1999.

During the first years of the appearance of the varroa mite it was supposed that the mite mainly damages the adult bees. But adult bees, however, succeed quite well in spite of being mite infested. The initial damage by the varroa mite takes place in the brood. One reproducing varroa mite in the brood cell is enough to severely influence the future life of the emerging bee. From a brood cell parasited by several mites fully developed bees cannot emerge. The clinical symptoms of dying brood resemble those of European foulbrood. Virus infection can worsen the damaging effect of the varroa mites.

During the season there is a steady increase in infestation rate of both brood and adults, and consequently, in the number of damaged bees. Varroa control exclusively by chemical preparations can only be successful at low infestations or if treatment early in the season is possible. It is necessary to take measures against the varroa during the whole year. This includes the cutting of drone brood, forming of nucleus and chemical treatments. In the middle and at the end of the season it is necessary to control the natural mite mortality. The number of mites in the debris gives an indication of the urgency of chemical treatments.

Under normal conditions, controlling parasites is not problematic. Generally, parasites do not kill their hosts because this would also endanger their own existence. From an evolutionary standpoint, the host-parasite-relations between varroa and the Western honey bee is not yet established because the transfer to the new host from the Asian honey bee took place quite recently.

For parasites that do endanger host survival control strategies are necessary. For the development of a control strategy it is of essential importance where, when and how the parasite damages its host. This paper comments on how strategies for varroa mite control in honey bee colonies should be developed.

During the first years of the appearance of the varroa mite it was supposed that the mite mainly damages the adult bees. But adult bees actually succeed quite well in spite of being mite infested. The loss of haemolymph (bee blood) is limited. At each sucking the mite hurts the pellicle between the abdominal rings (intersegmental membrane). As the mite can only absorb small quantities it hurts the bee several times within a comparatively short period, thus provoking the invasion of disease agents (fungi and bacteria). Moreover, viruses get into the bee corps via the saliva of the mite. Partly because of a well functioning endogenous immune defense system, the mite infestation and the loss of blood caused by a parasitation of up to two mites does not cause acute danger for the adult bee. Only if several mites attack the bee at the same time severe damages can be observed. Such a severe infestation of individual adult bees, however, only occurs close to colony collapse when mite populations have grown large enough to damage both brood and adults.

The initial damage by the mite takes place in the brood. One reproducing varroa mite in the brood cell is enough to severely influence the future life of the emerging bee. The life span of bees parasitised as pupae is shortened. Activities like hive hygiene, processing of food reserves and guarding services are also affected negatively. Parasitised brood produces bees that are not fully integrated as a productive member in the system of work division, which is of essential importance to the bee colony. Badly developed brood-food glands (hypopharyngeal glands) impede a long-term activity as nurse bee and the bee becomes a foraging bee at an earlier age.

From a brood cell parasitised by several mites fully developed bees cannot emerge. When the damage is visible, their life span is shortened to such an extent that they die already within a few days. This effect is, of course, even more pronounced for brood parasitised by a larger number of mites. Such bees often show deformations of wings and of the abdomen. Severe infestations of the brood often result in mortality already within the cell. The clinical symptoms of dying brood resemble those of European foulbrood. Especially the paralysis viruses (Acute Paralysis Virus and Slow Paralysis Virus) can worsen the damaging effect of the varroa mites. Viruses can provoke changes in bee behaviour similar to those caused by varroa mites. The life span of bees into which viruses are injected during pupae-stage in their development is shortened and such bees only work as nurse bees for a short time. With severe virus infections the brood dies showing symptoms of European Foulbrood. Deformations of the wings and the abdomen may also be caused by the deformed wing virus (DWV). As this virus only causes severe symptoms in a colony simultaneously infested by mites it was not considered important for practice before the introduction of varroa mites. It is generally valid that viruses strengthen the damaging effect of mites.

In certain years, viruses only appear in rare cases for reasons unknown up to now. When virus infections associated with the mite are not prevalent, bee colonies can survive populations of several thousand mites without problems. If, however, mites and viruses attack the bees, smaller number of mites may be enough to lead to colony collapse. Therefore, it is useful to recognize when an infestation by mites increases as early as possible and if there is a simultaneous virus infection. The mite infestation level can easily be observed by the beekeeper, the virus detection needs the help of a laboratory.

The mite reproduction possibilities increases with the amount of brood rearing and the duration of the brood rearing period. But also the bee density in a region and the subsequent pressure of infestation from colonies where varroa mite populations are not controlled finally decide the needs for mite control and the number of mites in a colony. Since the mite mortality, including the winter period, does not outweigh the mite reproduction capacity in any European climate where bees occur, the number of mites in the bee colony increases if left uncontrolled.

Under average conditions, in Central Europe, the mite infestation increases during the second half of the year. The more the number of brood decreases the more the number of brood cells infested by the mites increases. Thus, there is a steadily increasing number of damaged bees. In August/September the winter bees are reared. Only if the infestation of this brood is low the colony will manage to winter well and expand the following spring. Varroa control measures in July/August serve the need of getting fully developed winter bees. If varroa control is made only in September or later the proportion of damaged winter bees depends on the number of infested brood cells. If the treatment is late, many winter bees may be damaged and colony survival over winter will decrease.

Consequently, the following principle is valid: Every control measure of mites done before July/August is for the benefit of the winter bees and a later treatment helps the summer bees the following year.

Varroa control exclusively by chemical preparations can only be successful at low infestations or if treatment is possible early in the season. Chemical preparations usually produce residues in bee products. They should not be used before and during honey production. In highly infested colonies the bees are already severely damaged by the mites and the viruses, before a first chemical treatment is possible. In some years, even preparations with 100% effectiveness are not sufficient to save the colonies because the damages caused by the virus infections are not removed by removing the mites.

1) At the beginning of the season measures have to be taken to secure that the bee colonies enter the following nectar flow phase with the lowest infestation possible. As the application of chemical preparations for control immediately before a foraging period is not recommended, biotechnical methods, such as cutting out of drone brood and forming of nuclei, have to be used.

2) In the summer, the natural mite mortality is investigated in colonies equipped with net screen bottoms. The number of mites in the debris indicates the urgency of chemical treatments.

3) In the late season it is vital to control the natural mite mortality again. This is true for all methods of treatment irrespective of how effective they may be. Through reinvasion from collapsing colonies many mites may have penetrated into the colonies already treated or the chemical used for control may not be effective any longer because the mites may have developed resistance to the chemical used. The success of treatments should also be verified at this time. This is especially important if new method have been applied for the first time or if decreased efficacy of the preparation can be suspected due to resistance of the mites.

The control concepts can be adapted in many ways to the management method used, the hive system, and the requirements of the beekeeper. This is valid for the biotechnical methods as well as for the chemical alternatives. The method of cutting out drone brood depends very much on the hive type and the rhythm of operating the hives. Also the forming of nuclei can be made in different ways and at different times depending on the season and the general beekeeping system used. Chemical preparations are chosen mainly on the basis of their efficacy, where efficacy also in the sealed brood cells should be considered. Preparations not effective in sealed brood cells either require long exposure in the colonies, or have to be used in brood less colonies to be effective. So far, only formic acid treatments have been shown to be effective on mites in sealed brood cells. Another important factor for selection of chemicals for mite treatment is the danger of residues. Of course, only products registered for use in bee hives should be used for mite control and the label instructions should always be followed carefully.

At last, the hive system and the management method are again important criteria for what methods of mite control that are most convenient to adopt. Also personal preferences and general attitudes regarding chemicals and residues in bee hive products will be important criteria for individual choices of mite control strategies. For the practice it should be noted that it is not which chemical that is used which decides final success in mite control, but how a complete treatment concept is applied and adapted to local circumstances.

In order to make oxalic acid (OA) use against varroa mites easy and quick, a new administration technique was set up in 1996. According to this method an OA acidified sugar syrup (dehydrated OA, sucrose and water in ratio (weight) 1:10:10) is trickled by a syringe onto the top bars and in the space between combs with a dose of 5 ml/Dadant-Blatt comb populated by the bees. However, since OA and sucrose concentrations and doses were fixed empirically, some research to optimise this method was needed.

Specific trials were carried in various European countries (Finland, Germany, Italy, Norway, Sweden and Switzerland) to test different combinations between OA (0%, 2,1%, 3,2% and 4,2%) and sucrose (0%, 30%, 60% and 70%) concentrations. It was clear that sugar solutions containing 4,2% OA is the most effective: when administered according to the protocol it yielded an average mite mortality ranging between 90,3% and 97,8%. Nonetheless, in more limited trials the 3,2% option gave similar results. According to these trials, 2,1% OA concentration do not yield sufficient mite mortality for normal mite infestations.

Sucrose seems to be needed in the solution, since non-sugar options usually gave poor effectiveness. However, 30% sugar concentration could be enough since little (if any) differences were detected in comparison to 60% options for a given OA concentration.

With few exceptions, bee mortality at the hive entrances was normal after OA administration. No increase in winter mortality of treated colonies could be clearly related to the treatments. However, in mid-European trials some OA concentrations have demonstrated colony weakening as a result of OA treatments with an effect also registered the following spring. These negative effects seemed to disappear at 2,1% OA concentration. Italian observations, that need to be further confirmed, also indicate some negative effect on spring build up of colonies when 4,2% OA is used in high sugar concentrations (60% and 70%).

In The Netherlands, the effect of variations in the trickled amount was tested. A solution having 3,6% OA and 60% sucrose concentrations was used; one group of colonies received 2,9 ml/populated comb (corresponds to the same volume/surface ratio of 5 ml/Dadant-Blatt frame for the Dutch frame) and another one only 2,5 ml. The larger volume was more effective (92%), but was less well tolerated by the bees.

From colonies belonging to an OA treated and an untreated group, pupae were collected before administration of OA by trickling. Fifteen days after the treatment, newly emerged adults and pupae were sampled again from the same colonies. The Glutathione S-transferase (GST) activity was determined in the laboratory in each sample. Statistical comparisons could not demonstrate any significant difference in GST activity related to treatments. According to these findings it seems that an OA trickling at normal dose does not compromise the bee digestive system or weaken the detoxifying activity against potentially harmful substances.

In two experiments applying OA solutions by trickling and spraying were compared to the corresponding solutions based on potassium oxalate for their acaricide effect. The solutions compared in each trial differed in their pH, being very acid (OA) or neutral (potassium oxalate). The mite mortality was very low in the groups receiving the neutral solutions and much higher in the acid treated groups.

From these experiments, it seems that the acidity is responsible for the OA activity against the mites and not the oxalate ions in the solution; however the sensitivity to the acidity still remains unexplained.

The effect from oxalic acid (OA) on varroa mites has been known for several years. Past experiments have included various administration methods toTholonies: vaporisation (Popov et al., 1989); fumigation (with formic acid development during heating) (Okada & Nekane, 1987); spraying weak solutions onto adult bees (Radetzki et al., 1994; Nanetti et al.; Imdorf et al., 1995), and trickling OA acidified sugar syrups into the colonies (Nanetti & Stradi, 1997). Some of these techniques showed a very high efficacy, although different experimental conditions and methods make direct comparisons difficult

In some European countries spraying bees in broodless colonies with a 2-3% OA in water solution is commonly used amongst beekeepers. This technique is very effective but, since every comb of bees has to be sprayed on both sides, it is usually considered too labour-intensive and time-consuming (and therefore uneconomical) by large-scale beekeepers.

Attempts to make application easier led to a new administration technique being developed. An OA acidified sugar syrup made up of dihydrated OA, sucrose and distilled water, 1:10:10 by weight, (OA concentration 4,2%, sucrose concentration 60%) is trickled by syringe into broodless colonies. A 5 e ml dose for each Dadant-Blatt comb (dm (3,00*4,35) = dm2 13,05/side) occupied by adult bees is used. This is eqivalent to 0.38 ml solution per square dm covered by bees. Trials in Italy showed a high acaricidal activity (96,8% and 96,1% in 1996 and 1997 respectively) without any ill effects in test colonies, even at very low temperatures (Nanetti & Stradi, 1996 and unpublished data). In 1996 an efficacy of 89,6% was recorded using a solution with the same sugar strength but an OA concentration of 2,2%. On the basis of these results, and on the results of trials performed by beekeeper organisations, many Italian hobbyist and professional beekeepers now use this method in autumn treatments against varroosis.

However, in some instances, bees show poor tolerability to OA trickled into the hive. Colonies receiving an overdose (i.e., excessive amounts, short-term repeated administrations, or excessive OA concentrations) can be weak at the end of the winter or sometimes fail to over-winter at all. In these cases no abnormal bee mortality is usually seen at the hive entrance. However, according to both published (Charrière et al., 1998) and unpublished data, some trials have also found that bees do not tolerate the trickling treatment very well at normal dosages.

An OA specialist team from within the CA3686 group developed a series of experiments to improve our understanding of the treatment, with special reference to the trickling technique. The findings so far are summarised below, together with the results of independent, but related, experiments by researchers belonging to the group.

Sugar and OA concentrations for the solution used in Italy were chosen empirically. This experiment aimed to answer the following questions:

- can better results be obtained using different OA and/or sucrose concentrations?

A two-stage experiment was set up. Nine solutions were screened (i.e., every combination of 0%, 2,1% and 4,2% OA concentration and 0%, 30% and 60% sugar concentration) in Nordic countries, where early brood interruption is expected. The most effective combinations were then tested in Southern European areas, where the brood rearing season is longer. However, some extra solutions not included in the Nordic tests were also tested at the second stage. This procedure allowed both stages to be conducted in the same autumn-winter period.

The dose administered varied according to colony strength and, in most cases, to comb size. It was trickled into the gap between two populated combs and onto the respective top bars. Efficacy was calculated as a ratio between the mite mortality due to the treatment and the total infestation level based on control treatments with acaricides. Bee tolerability was assessed by checking mortality outside the hive and by comparing the colony conditions before the treatment, at the end of the winter, and at the beginning of the honey flow. These evaluations were performed by the Liebefeld method.

Table 1 summarises the efficacy recorded in the first stage of the study. Column one of the Finnish results refers to doses calculated on the basis of 100% populated combs; this generally led to under-dosing the treatment and therefore to lower effectiveness.

The highest efficacy was achieved with sugar solutions having the highest OA concentration. The actual values are very close to Italian results. A 2,1% OA solution generally produced a lower efficacy and more variable results.

Solutions without sugar resulted in poor and variable efficacy, even at a 4,2% OA concentration. Also, no clear differences in acaricidal activity were observed between the 30% and 60% sucrose solutions for a given OA concentration, in the same apiary.

Table 2 summarises the average efficacy recorded in the second stage of the study. Again the combination 4,2% OA and 60% sucrose resulted in a remarkably high acaricidal activity. However, the treatments in Italy resulted in a slightly lower effectiveness than that recorded in previous Italian trials which might be due to the low temperatures during the treatment period in this experiment. It is possible that the reduced efficacy in Finnish trials are related to the much lower temperature during treatment in Finland. The effects from temperature during treatments need further studies. Nevertheless, the 30% and 60% sugar solutions for a given OA concentration still yielded similar results and much better efficacy than using water solution. However, in Switzerland also the sugarless solution gave a noticeably high efficacy. The weakest OA solution (2,1%) resulted in less than 90% efficacy, but 3,2% OA approached the effect of the strongest one.

Although the results for acaricidal activity were substantially consistent, some important differences in bee tolerability were observed. The end of winter and the spring colony development measurements carried out in Scandinavia did not demonstrate any detrimental effect on the colonies caused by any combination of OA or sugar concentration. However some bee mortality was observed outside hives in the days immediately after the treatment in a few cases. In contrast, evaluations made in Switzerland and Germany showed a tendency to higher bee losses during winter with the 4,2% OA sugar solution. Lower concentrations resulted in a better over-wintering and spring development.

In Italy, when the treatment was administrated in a cold environment (mountainous), sugar concentration (30% and 60%) did not seem to affect either over-wintering or spring development. However, in foggy conditions the spring measurements indicate that there may be some long-term reduction in colony strength when 4.2% OA solution is used with high sugar concentrations (60% and 70%).

The colony losses recorded during winter in the different countries were no higher than would normally be expected. None could be clearly related to treatment.

The 4,2% OA solution in 60% sucrose showed a remarkable effectiveness against varroa mites in broodless colonies treated in the autumn/winter using the trickling method. Although in many cases the bees seem to tolerate the treatment well, in some conditions this OA concentration caused poorer over-wintering or a slower spring development of colonies. Short-term mortality was not seen outside the hive.

Tolerability is better with 3,2% OA. According to the mid-European experiments the differences in efficacy of this concentration compared with 4,2% OA is of little consequence for the beekeeper. A 2,1% OA concentration, however, reduces the risks for the colony but its lower efficacy means that it can only be used when the infestation level is rather low.

Sugarless solutions have poor efficacy and are not practicable, but a reduction of sugar concentration from 60% to 30% seems to have no significant influence on acaricidal activity. This also seems to be true for OA concentrations of less than 4,2%. There is also a possibility that a lower sucrose concentration enhances the ability of bees to tolerate the treatment solution, although no conclusive evidence of this emerged from the trials. Further investigation is needed.

The possibility that high environmental humidity levels reduce the bee tolerability should also be further studied.

We need to verify the optimal amount of solution that must be given to the colonies. A trial was conducted in the Netherlands with this in mind. One solution of 3,6% OA in 60% sucrose was used. It was trickled into colonies according to the method described above, but in variable amounts. Some colonies received 2,9 ml/comb (this dose corresponds to 5 ml/Dadant-Blatt comb because Dutch combs are smaller) and other ones were given 2,5 ml/populated comb. A group of untreated colonies was used as a control.

Efficacy was calculated as the ratio of the mite mortality due to the treatment to the total infestation level. Colony strength was evaluated by shaking and weighing the bees before treatment and then again in the spring.

The larger volume was 92% effective, in line with the values found in the previous trials using 3,2% and 4,2% OA sugar solutions. However, the smaller volume resulted in a remarkably lower effectiveness (80%). A reduced bee population was found in both treated and untreated colonies in spring compared to pre-treatment. The colony weakening seemed to be related to the amount of solution given to the colonies. The reduction of the bee population was 72%, 58% and 41% for colonies treated with 2,9 or 2,5 ml/comb (same amount of active ingredient) or in untreated controls respectively. This experiment underlines the need to establish the optimal dose that will be both effective and safe for bees. Further trials are still clearly needed.

The tissues of the honey bee gut have glutathione S-transferase (GST) activity. This important group of enzymes acts as a detoxifying system against potentially harmful substances with which bees may come into contact. A lowering of GST activity could make bees more vulnerable to environmental toxic substances.

The following experiment (Brødsgaard et al., 1999) was carried out to discover if OA-sugar solution ingestion reduces GST activity in bees:

Pupae were collected from colonies immediately before treatment by OA trickling and from untreated controls. Fifteen days after the treatment newly emerged adults and pupae were taken from the same colonies. The average GST activity in these samples is summarised in table 3.

The statistical tests performed on the data sets for both pupae and adults did not show any significant difference in GST activity attributable to treatment (comparisons between groups and within each group). However the treatments were performed using normal doses. It is still possible that some detrimental effect on GST activity may occur in case of overdosing. More research is needed to fully understand the treatment-related negative effects, with special consideration being given to colony weakening that can result in certain circumstances.

The pH/concentration curve is much lower for OA solutions than for the majority of organic acids – even at low concentrations OA, the acidity is very high. The theoretical pH values for the solution used for spraying onto bees (2,1%) and for the solution used for trickling (4,2%) are about 1 and 0.9 respectively. Besides, both solutions have a noticeable efficacy against varroa mites.

To understand whether the OA effect on mites is related to the chemical and toxicological features of oxalate ions coming from acid dissociation in water, or to the high acidity of the solution, two experiments were carried out on broodless colonies. In the first experiment the trickling method was used, whereas spraying was used in the second. In each case one group of colonies was treated by an OA solution (acid) and a second group received a potassium oxalate solution (neutral) both having the same molarity. Efficacy was calculated for each colony as the ratio between the mite mortality due to the treatment and the infestation level.

In both experiments the mite mortality was very low in the groups receiving the neutral solution, but in the other groups a very high average efficacy of about 90% was recorded (table 4). It seems that acidity is responsible for the OA action against varroa mites but the reasons for this are not known. The poorer efficacy of other organic acids (e.g., lactic and citric acids) could be due to their lower dissociation constant.

TABLE 1. Average effectiveness recorded in the 1^st stage of solution optimisation experiment.

Apiary	Sugar	OA	FINLAND		SWEDEN	NORWAY
1	0%	0,0%	5,5		5.0	5,4	2,0
	0%	2,1%	39,1		22.3	39,2	36,7
	0%	4,2%	48,8		50.9	86,8	75,0
	0%	1,0%			11.0
2	30%	0,0%	5,3		4,4	4,1	2,0
	30%	2,1%	35,9		39.1	62,4	63,0
	30%	4,2%	92,9		94,8	91,3	84,0
3	60%	0,0%	8,2	14,9	4.2	9,3	28,4	2,7
	60%	2,1%	10,8	81,5	68.2	85,5	82,5	32,2
	60%	4,2%	79,7	93,2	96.1	96,5	95,1	93,8
5	0%	2,1%	5,4			65,2	69,3
	30%	2,1%	6,0			93,5	88,9
	60%	2,1%	15,4			92,5	82,9
	0%	0,0%					5,3
	60%	0,0%				3,6
6	0%	4,2%	29,0
	30%	4,2%	76,4			97,2	92,7
	60%	4,2%	83,9			97,8	94,3
	60%	0,0%				7,6	4,4

Sugar	OA	SWITZERLAND	GERMANY	ITALY
0%	4,2%	92,5
30%	4,2%			89,6
60%	0,0%	3,5	3,4	3,2
60%	2,1%	86,7	84,8
60%	3,2%	98,6	92,2
60%	4,2%	97,5	94,3	90,3

TABLE 3. Average Glutathione S-transferase activity in honey bees before and after OA trickling (Brødsgaard et al., 1999).

Essential oils and essential oil components offer an attractive alternative to synthetic acaricides for the control of Varroa jacobsoni. They are generally inexpensive and most pose few health risks. Terpenes (mainly monoterpenes) are the main components of essential oils, comprising about 90% of the total. More than 150 essential oils and components of essential oils have been evaluated in laboratory screening tests. Very few of them, however, have proven successful when tested in field trials. Thymol and thymol blended with essential oils or essential oil components offer a promising exception. Mite mortality obtained with these formulations typically exceeds 90% and often approaches 100%. In addition, residues in honey are low, even after long-term treatments. The exact conditions under which these formulations will yield reliable and effective control, however, have only been determined for certain European regions. Based on the available studies, relying solely on a single treatment with an essential oil or essential oil component is generally not sufficient to maintain mite populations below the economic injury level. Therefore, efforts are necessary to optimise the use of these substances and to incorporate them, along with other measures for limiting mite populations, into an integrated pest management strategy for control of Varroa jacobsoni.

Colonies in temperate regions must be treated once or twice a year against Varroa jacobsoni to maintain their populations below economic injury levels. During the last 10 years, the pyrethroids have been the primary source of insecticides used to control V. jacobsoni. Recently, mites in parts of Europe and North America have developed resistance to pyrethroids. The widespread use of synthetic lipophilic acaricides has lead to the accumulation of residues in beeswax, propolis and to a much lesser degree, in honey.

The development of acaricide resistance in V. jacobsoni populations and the spectre of the contamination of hive products provide considerable incentive to develop new treatment strategies that minimise the potential for acaricide resistance and the accumulation of residues. Since V. jacobsoni was introduced to Europe, intensive efforts have been made to develop alternative chemical control measures based on formic, lactic and oxalic acids combined with biotechnical measures.

It is well known that many essential oils and their components exhibit acaricidal activity. Before V. jacobsoni was a world wide pest, different components of essential oils were tested for their activity against Acarapis woodi. Methyl salicylate and menthol proved to be toxic to the tracheal mite. In the last 15 years, research has shown that several essential oils and individual compounds of essential oils also have a high acaricidal activity against Varroa jacobsoni.

In extensive screening tests, many oils show significant acaricidal activity. Some oils are repellent to V. jacobsoni, others are attractive, and some cause mite mortality. However, of more than 150 essential oils and components of oils tested, only very few have proven effective when applied in hives in field trials. This is most probably due to the fact, that the screenings tests used were incapable of predicting the acaricidal effect under field conditions. Difficulty in obtaining standardised essential oils also affects treatment predictability. Only a combination of wintergreen oil and thermal treatment, an aerosol treatment of a thyme-sage oil mixture, and the passive evaporation of thymol, oregano oil and marjoram oil in combination with diluted formic acid have been used successfully for mite control. For different reasons, however, none of these treatments have been widely adopted by beekeepers, with the exception of thymol. Indeed, thymol and thymol blends are widely used to control V. jacobsoni in Europe and in most cases their varroacidal efficacy is greater than 90 % (Fig. 1 and 2). Different thymol containing products are available on the market.

Because of insufficient predicting capacity of the screening tests used until now, we devised an assay, in which the dose-response relationship of an airborne acaricide and the corresponding mite and bee mortalities can be assayed under laboratory conditions. Using this technique, a high mite toxicity, combined with good bee tolerance, were demonstrated, besides thymol, for the following components of essential oils: camphene, camphor, p-cymene, eugenol, isopinocamphon (ysop oil), menthol and a -thujone. Identifying compounds with acceptable acaricidal activity but with low toxicity to honey bees is essential for providing candidate compounds for field trials. After finding suitable substances under laboratory conditions we will measure the air concentration under field conditions to test their efficacy in a bee colony. This procedure can serve as a powerful screening technique because it guides subsequent field research into the most productive avenues. The development of effective delivery systems for essential oils remains one of the greatest obstacles to their implementation as mainstream control measures. Highly volatile substances like camphor are difficult to use, but formulations retarding the evaporation rate, e.g. special gels, might overcome this difficulty. Products with mixtures of different components with different modes of action, might also provide effective solutions. For example, substances that disrupt the mite’s host finding behaviour may be effective in conjunction with substances that kill mites.

Residues pose another challenge to the use of essential oils. Most essential oils are mixtures of more than 50 components. Depending on the individual partition coefficients of the constituents, residues in honey and wax are to be expected. Residues in honey can lead to adverse effects on taste, while residues in wax can render it unsuitable for some applications. Quantitative residue analyses are required for product registration. The complex nature of many essential oils, combined with the fact that many essential oil components are naturally occurring in honey, makes such residue analysis difficult. Thus, the successful development of products employing essential oils can be extremely difficult unless the particular essential oil has been granted an exemption from existing regulations on maximum residue limits. In the EU, thymol, menthol and camphor have this status. The use of individual components of essential oils makes residue analysis much easier and limits the potential for producing off-flavour honey. Long-term studies have demonstrated that when used properly, residues of thymol in honey remain at low and safe levels (Tab.4).

Based on the available studies, relying solely on one treatment per bee season with essential oils or essential oil components can not be recommended as an effective and reliable method to maintain mite populations below the economic injury level. The challenge for future research is to optimise the use of essential oils and essential oil components and to incorporate the resulting products along with other measures for limiting mite populations such as cutting out of drone brood, trapping combs, formation of nucleus colonies or the use of organic acids into an integrated pest management strategy for the control of V. jacobsoni. Adapting these strategies to local climatic conditions, to diverse apicultural management practices and to beekeeping operations of varying sizes pose additional and significant challenges. Finally, resistance to essential oils may eventually develop, as it has with synthetic pesticides. Consideration must be given to the development of resistance management plans to maximise the useful life span of effective acaricides and delivery systems once they are developed.

The results, reported in the present manuscript, are presented in detail in the review cited below.

Table 2 - Treatment of V. jacobsoni with blends of thymol, eucalyptol, camphor and menthol (N.C. = non commercial)

Type of thymol treatment	average mg/ kg	Min-Max mg/kg
Thymol frame a whole year use in Switzerland, 1997 ( n = 22)	0.33	£ 0.02-0.83
Thymol frame a whole year use in Switzerland, 1998 (n=34)	0.40	0.11-1.06
Thymol frame use outside the honey flow period in Switzerland 1998 (n=10)	0.17	£ 0.02-0.32
Thymol frame a whole year use in Germany, Wallner 1997 (n = 19)	0.63	0.07-2.0
Api Life VAR 8 weeks treatment in autumn, 1 to 5 use (n=28)	0.16	£ 0.02-0.48
Lime honey (Guyot et al. 1998)	0.08	0.02-0.16
Thymol concentration affecting honey taste	1.1-1.3
Maximum residue limit for Switzerland	0.8

For the control of varroa mites, formic acid has been applied by individual beekeepers for more than 20 years (2;25-28;35;47;52). Use of formic acid (FA) represents an essential part of an integrated management for the control of Varroosis.

Formic acid has to evaporate within the beehive and it kills the mites by respiratory inhibition. To give full effect, high concentrations of FA in the air have to be present for several hours or even days. Short-term treatments need to reach higher concentrations for full efficacy, compared to long-term treatments. FA is the only chemical treatment which kills the mites both on the bees and within sealed brood cells (1;6;10;17).

FA can be used successfully only when the ambient temperature is >12°C and sealed brood is present in the colonies. Efficacy and side effects depend on the dynamic of the evaporation of FA within the hive. Evaporation is influenced by

Because the large number of variables involved, standardisation of FA evaporation is difficult and results in large variations in efficacy among the application devices available on the market. In general, two types of application systems can be distinguished; short-term and long-term treatments. Depending on the method of treatment, a high concentration of formic acid for a few hours or a low concentration for several days is achieved in the air of the bee hive, with the respective system (14).

During the short-term treatment small amounts of formic acid evaporate relatively uncontrolled within 6 to 10 hours. At the beginning of the treatment the formic acid concentration in the hive air increases rapidly (Fig 1). Six hours later most of the acid has already evaporated. For proper use, the application method has to be adjusted to the ambient temperature and to the type of the hive used. When FA is applied from above 60% formic acid is used, whereas 85% may be necessary if the acid is applied from below (Tab 1). In Switzerland, the treatment in two blocks of two to three applications within a week in August after the honey harvest and in the end of September has proved to be efficient. The treatment efficacy obtained under these circumstances is approximately 90 to 95% (7;9;15;16;23;24;31;35;38;49;53).

The efficacy can be controlled by counting the natural mite fall starting two weeks after the last treatment by use of a bottom board with a metal screen covering the entire hive bottom. It is sufficient to count the mites once a week. If the natural mite fall is higher than one varroa mite per day, another treatment with oxalic, lactic acid or another acaricide is recommended (22). Six years’

Concentration of formic acid in the air of the hive during the application of a short- and a long-term treatment (1ppm = 1.91 mg/m³)

Experience with this method has shown, that follow-up treatments are necessary only after reinvasion in October (21). In order to avoid bee and queen losses, the indications below concerning temperature and application must be followed and the hive entrance has to be entirely open. Feeding at the time of the treatment can reduce side effects (but also efficacy).

Practical experience has shown, that short-term treatments alone often do not reduce the varroa population sufficiently to prevent a detrimental increase in the mite infestation level the next summer. To successfully rely on short-term FA treatments for mite control, it is necessary to reduce the increase of the varroa population also in the spring by removal of 2 - 3 drone brood combs or removal of sealed worker brood by forming nuclei.

Most short-term treatment methods do not require expensive devices or additional hive material and can be easily applied immediately after the last honey yield. Short-term treatments are an important module within an integrated concept to reduce mite infestation levels as early as possible during the season to prevent collapse of bee colonies and reinvasion of mites.

In type 1 dispenser the acid concentration within the container keeps stable (no exchange with air humidity) and the amount of evaporated FA can be easily controlled.

Compared to several short-term treatments, the amount of work involved using long-term treatments is reduced considerably. Various types of evaporation devices for long-term treatments are available on the market. Also after long-term treatments an additional autumn/winter treatment with a systemic acaricide is recommended (depending on diagnosis of natural mite mortality) and/or removal of drone brood the following spring. By combining different control methods, it is not essential to achieve the highest possible efficacy using the FA treatment(s). A lower intensity of FA treatment can reduce the risk of queen losses considerably. Depending on the degree of infestation, one or two long-term treatments with formic acid are necessary. Under Swiss conditions, two long-term treatments are needed if natural mite fall is higher than 10 mites per day at the beginning of August. The first treatment over a one week period should be done immediately after the honey harvest. Depending on climatic region (ambient temperatures during the day should be > 15°C!) the second treatment of about two weeks should be done in mid-September. For colonies with a mite mortality of less than 10 mites per day, one FA treatment of two weeks in August/ September is sufficient (19).

Active substance	Formic acid – short-term treatment
Application	Passive evaporation from absorbent material
Period of treatment	Begin: After honey yield End: Depending on ambient temperature Recommendation in Switzerland: 1st block of treatment Ÿ beginning of August 2nd block of treatment Ÿ end of September (Duration of block of treatment app. 1 week)
Number of treatments	2-3 treatments per block
Temperature during the day	12-20 °C Ÿ treatment during the day 20-25 °C Ÿ treatment in the evening or morning > 25 °C Ÿ treatment early in the morning
Concentration	Treatment from the top Ÿ 60 % FA Treatment from the bottom Ÿ 60% - 85 % FA(depending on ambient temperature)
Dosages (depending of the size of the hive)		From the top From the bottom	1story (ml) 20-30 20-30	2 story (ml) 40-50 40-60
Absorbent material	Viscose sponge (slower evaporation) Soft fiber cardboard (quick evaporation) Soft fiber Pavatex plate (wood fiber)
Surface of evaporation	Approximately 15 x 20 cm
Control of treatment efficiency	Measuring of the natural mite mortality Beginning Ÿ 14 days after last treatment Duration Ÿ 2 weeks More than 1 varroa/ per day Ÿ additional treatment is recommended
Security measures by the application	Protective glasses, rubber gloves and water available

Applied after the honey harvest in autumn, the natural content of formic acid in the honey of next spring may under certain conditions increase slightly (20;42;51). This increase has no negative effect on the honey quality (5).

The efficacy of one FA treatment will range from about 60% to 80%. For two applications, the efficacy can increase to 90-95%. The efficacy has been extensively tested for the Krämer Plate (8;29;30;39;50), the Nassenheider (3;4;7;36;43-45), the FAM-Liebefeld dispenser (12;13) and in recent times the Tellerverdunster (32-34;48) . In comparative tests in three apiaries (11) the efficacy of 5 dispensers were between 92 and 98% (two treatments). This efficacy can be regarded as sufficient to obtain control of the mite population when used in an integrated system (Tab 2). More dispensers such as the Universalverdunster (37) and the gel application (40) where an efficacy of more than 90% can be expected, has recently been developed.

Although the average effect of FA treatments may be acceptable, there are often colonies or apiaries with insufficient treatment results because of variations in efficacy. Such colonies or apiaries represent a risk for other colonies because of reinfestation.

An important point is that the application of FA has to be included intensively in the beekeeper extension work and that details of the method has to be adapted to the local conditions (climate, bee hive). It will not be possible to develop a general application guideline for all of Europe as we are confronted with a wide range of climatic conditions, hive types and different bee management system. Finally it seems that under Mediterranean conditions FA application is more difficult, probably because of the high and often varying temperature.

Tab 2 Efficacy of formic acid against Varroa jacobsoni after the application of 2 long-term treatments with 5 different dispensers.

	Treated group	Control group
Pupae (before treatment)	194,1	184,4
Pupae (15 days after the treatment)	207,0	191,5
Newly emerged adults (15 days after the treatment	226,2	211,5

Authors	year	thymol formulation	dosage	place of application	days of treatment	time of treatment	# of colo-nies	# of supers	# of apia-ries	type of hive	mean treatment efficacy %	mean treat-ment mitefall
Marchetti et al.	1984	powder in bag	4 x 15g	between combs	16	Oct./Nov.	10	1	1	Dadant	66.0	3229
Lodesani et al.	1990	powder	3 x 4.5/6g	dusted over combs	21	Oct./Nov.	38	1	2	Dadant	81.0	190
Frilli et al.	1991	powder	4 x 1g	on comb bars	8	Nov.	7	1	1	Dadant	95.0
Chiesa	1991	powder	5 x 0.5g/comb	on comb bars	8	Oct./Nov.	21	1	3	Dadant	96.8	1917
Liebig	1995	on compound	2 x 15g	on comb bars		Aug./Nov.		1		Zander
		on compound	2 x 30g	on comb bars		Aug./Nov.		2		Zander
Higes et al.	1996	powder	5 x 1g beeway	on comb bars	19	Feb.	4	1	1	Autocol.	97.8	977
Higes and Llorente	1997	powder	4 x 8g	petri on combs	28	Apr./May	4		1	Langstr.	97.6	1119
Flores et al.	1997	powder	2 x 10 g	petri on combs							97.0
		on compound	2 x 10g	on comb bars							95.0
Bollhalder	1998	on compound	2 x 15g	on comb bars	49	Aug./Oct.	22	1	4	CH	85.0-97.0

Authors	year	product	number of com-pounds	place of appli- cation	days of treat-ment	time of treatment	number of colonies	number of supers	number of apiaries	type of hive	mean treatment efficacy %	mean treatment mitefall
Contessi and Donati [22]	1985	Biovarroin	2 x 1	top	35	Nov./Dec.	2	1	1	Dadant	92.6	316
Tonelli [88]	1989	Api Life VAR	2 x 1	top		Nov./Dec.					93.8
Rickli et al. [80]	1991	Api Life VAR	2 x 1	top	38	Aug./Sep.	20	1	1	CH	96.4	986
		Api Life VAR	2 x 1	top	79	Aug./Oct.	20	1	1	CH	99.0	2453
Mutinelli et al. [unpbl.] data]	1991	Api Life VAR	2 x 1	below	40		13	1	1	Dadant	89.0	593
van der Steen [91]	1992	Api Life VAR	2 x1	top	42	Sep./Oct.	5		1		74.0
		N.C.+ camphor	2 x 1	top	42	Sep./Oct.	5		1		92.0
		N.C. - camphor	2 x 1	top	42	Sep./Oct.	5		1		88.0
Moosbeckhofer [76]	1993	Api Life VAR	2 x 1		29	Sep./Oct.	23	2	3	Zander	98.6	1400
Mutinelli et al. [77]	1993	Api Life VAR	2 x 1	top	49	Aug./Oct.	27	1	4	Dadant	68.7	4925
Liebig [59]	1993	Api Life VAR	2 x 1	top		Sep./Dec.	14	1	4	Zander	97.4	1276
		Api Life VAR	2 x 1	top		Sep./Dec.	26	2	4	Zander	63.9	1276
Schulz [84]	1993	Api Life VAR	2 x 1	top		Aug./Dec.	3	2	1	Zander	74.7
		Api Life VAR	2 x 2	top		Aug./Dec.	4	2	1	Zander	94.9
		Api Life VAR	2 x 3	top		Aug./Dec.	2	2	1	Zander	99.5
		Thymix	2 x 1 or 2	top		Sep./Dec.	77	1 or 2	7	Zander	94.8	3492
Imdorf et al. [50]	1994	Api Life VAR	2 x 1	top	56	Aug./Oct.	83	1	8	CH	97.7	602
Imdorf et al. [46]	1995	Api Life VAR	2 x 1	top	42-56	Aug./Oct.	19	1	1	Dadant	91.7	1078
Calderone and Spivak [15]	1995	N.C.	2 x 2	top	19	Nov.	8	2	2	Langstroth	96.7
Gregorc and Jelenc [35]	1996	Api Life VAR	2 x 1	top	30	Aug./Sep..	14	2	1	Alberti.	66.4
Loglio et al. [65]	1997	Api Life VAR	3 x _	top	21	Jul./Aug.	32	1	1	Dadant	72.6
Calderone [14]	1999	N.C.	2 x 1	top	32	Oct./Nov.	6	2	1	Langstroth	67.0

Advantages		Disadvantages
Useful in colonies with brood		Variation of efficacy (as exclusive treatment method mostly not sufficient)
No negative effects on honey quality		Minor brood damages have to be accepted
No resistance of the mite reported so far		Several treatments necessary
Important module for integrated concepts		To many methods on the market ("confusion")
15 years experience in Europe available		Requirement of beekeeper extension with adaptation to local conditions
Also toxic for tracheal mites (18;41;46;54).

		Efficacy (%)
Dispenser	Colony number	1st FA treatment	2nd FA treatment	Total FA
Apidea	14	59	89	96
Burmeister	14	42	87	92
FAM-Liebefeld	13	74	91	98
Krämer Plate	13	37	92	95
Wyna-Deluxe	10	75	85	96