Parasite-host interactions between the Varroa mite and the honey bee : a contribution to sustainable Varroa control

<h3>Introduction</h3><h4>Varroa mites as parasites of honey bees</h4><em>Varroa destructor</em> (Anderson & Trueman, 2000), is the most important pest of European races of the Western honey bee, <em>Apis mellifera</em> L., weakening bees and vectoring bee diseases (Matheson, 1993). Over the past decades it has spread all over the world and control measures are required to maintain healthy honey bee colonies.Originally, this mite only occurred in colonies of the Eastern honey bee, <em>Apis cerana</em> Fabr., in Asia. <em>Varroa destructor</em> was formerly known as <em>V. jacobsoni</em> Oud. (Anderson & Trueman, 2000). The <em>Varroa</em> mite was described in 1904 by Oudemans as a parasite of Eastern honey bees in Indonesia. Although the actual damage inflicted by the mite to the Eastern honey bee has never been determined, the <em>Varroa</em> mite is not considered to be a problem in colonies of its original host. However, <em>Varroa</em> turned into a serious pest of Western honey bees when beekeepers moved the Western honey bee into the area of distribution of the Eastern honey bee. The mite appeared to be a harmful parasite on its new host, but before this was realised it had already spread over the world through shipments of colonies and queens (De Jong et al., 1982; Matheson, 1993).<em>Varroa</em> mites may ruin Western honey bee colonies because parasitised bees suffer from malformations and a shortened life span (Beetsma et al., 1989). The <em>Varroa</em> mite feeds on both adult bees and brood, but reproduction is restricted to brood cells, which mites invade during the final larval developmental stage of the honey bee. Offspring is produced during the period that the immature bee develops in the capped brood cell and the mother and her progeny emerge together with the young bee. In addition to direct damage to bees through feeding, mites act as vectors of honey bee pathogens and increase the incidence of honey bee diseases (Ball, 1994). This threat of <em>Varroa</em> mites to beekeeping resulted in the development of acaricides and nowadays several effective acaricides are available which are applied world-wide (Koeniger & Fuchs, 1988; Ritter, 1990). However, the use of acaricides has important disadvantages. Acaricides contaminate bee products like honey and wax (De Greef, 1994) and thus the use of these acaricides is in conflict with the status of honey and wax as natural products. Another disadvantage is that mites have become resistant to these acaricides and this resistance is spreading world-wide, which implies the need for alternative ways of control.<h4>Towards sustainable Varroa control</h4>In this thesis, I present studies on biotechnical methods of <em>Varroa</em> control and studies on how variation in the honey bee's susceptibility to <em>Varroa</em> affects the mite population growth. In theory, biotechnical control methods in which mites are trapped in brood cells and removed from the colony, so-called trap-comb methods, are simple. In practice, however, these methods may become complicated because timing of application needs to be integrated in other activities of the beekeeper, such as swarm prevention. In addition, application of these methods is usually labour intensive. Effective trap-comb methods are available, but reduction of labour intensity is still needed. Much research is therefore directed to breed honey bees that are less susceptible to <em>Varroa</em> mites (Woyke, 1989; Büchler, 1994; Moritz, 1994). In this field, I investigated whether reduced developmental time of bee brood and attractiveness of bee brood to mites are suitable traits for selection aiming at reduced susceptibility of honey bees to <em>Varroa</em> mites. If less susceptible honey bees are available, the high effectiveness of control methods needed for successful control may be relaxed. This in turn may allow simplification of biotechnical control methods. The aim of my thesis is to develop acaricide-free beekeeping by using alternative methods for effective control of <em>Varroa</em> .<h4>Objectives and research questions</h4>Applying knowledge on invasion behaviour in the development of biotechnical control methods and population modellingThe parasite-host interactions between the mite and the honey bee have been intensively studied, because such knowledge may lead to new ways of control. In earlier work, I collaborated with Beetsma and Boot (1995) to study invasion behaviour of mites into brood cells. <em>Varroa</em> mites survive on adult bees, but reproduction is restricted to the capped brood cell (Ifantidis & Rosenkranz, 1988). The rate of brood cell invasion defines the distribution of mites over bees and brood and, therefore, the population dynamics of the mite. The rate of invasion appeared to depend mainly on the ratio of brood cells that are being capped per bee in the colony, as reviewed in Chapter 1. In this thesis I applied this knowledge to design control methods that are based on trapping mites in bee brood. I investigated if it is possible to predict the effectiveness of trap-comb methods using a model based on the calculated invasion rate of the mites in brood cells from the ratio of capped brood cells per bee (Chapters 2&3). Using this model, concepts of trap-comb-methods were evaluated (Chapter 4). I also applied knowledge on invasion behaviour to gain more insight in the mite's population dynamics in general (Chapter 5).<h4>Towards less susceptible honey bees</h4>Differential reproduction of mites in both host-species, <em>A. cerana</em> and <em>A. mellifera</em> , seems to be a key factor in susceptibility of honey bees to <em>Varroa</em> (Büchler, 1994; Rosenkranz & Engels, 1994). In European <em>A. mellifera</em> colonies mites reproduce in both worker and drone brood and mite numbers increase rapidly. In colonies of its original host, <em>A. cerana</em> , mites invade both types of brood cells but refrain from reproducing in worker cells (Boot et al., 1997). Thus, in <em>A. cerana</em> mite numbers can only increase when drones are being reared. In African and africanised <em>A. mellifera</em> races a high percentage of mites that invade worker brood also refrain from reproducing (Camazine, 1986; Ritter, 1993). Therefore, like <em>A. cerana,</em> African and africanised honey bees are less susceptible to <em>Varroa</em> . I studied whether refraining from reproduction in worker brood is due to a trait of the honey bee or due to a trait of the mite (Chapter 6). By transferring <em>Varroa</em> mites originating from <em>A. mellifera</em> colonies to <em>A. cerana</em> worker brood and vice versa there appeared to be two distinct mite populations with a different reproductive strategy. Mites originating from <em>A. mellifera</em> reproduced in worker brood in both species of honey bee, whereas mites originating from <em>A. cerana</em> reproduced in drone brood only. Later, genetic studies of <em>Varroa</em> mites (Anderson & Trueman, 2000) made clear that the two populations in fact belong to different species. The mites that parasitise Western honey bees originate from Korea and Japan and were erroneously called <em>V. jacobsoni</em> and have been recently named <em>V. destructor</em> (Anderson & Trueman, 2000).Selection for honey bee traits that reduce reproductive success in worker brood is reminiscent of the situation we in the original host-parasite relationship where mites reproduce exclusively in drone brood. I studied honey bee traits that may play a role in the reproductive success of <em>Varroa</em> mites in worker brood: the duration of the capped brood stage and attractiveness of the brood cells. A short duration of the capped brood stage will limit the development of nymphs (Chapter 7). Reduced attractiveness will decrease the rate of invasion and hence the rate of reproduction (Chapter 8).<h3>Summary</h3><h4>Structure of the thesis</h4>The chapters in this thesis are articles in which a separate part of the work is introduced and results are presented and discussed. The first six chapters have been published in periodicals and the final two chapters are submitted for publication.<h4>Invasion behaviour of Varroa mites: from bees into brood cells (Chapter 1)</h4><em>Varroa</em> mites may invade worker or drone brood cells when worker bees bring them into close contact with these cells. The attractive period of drone brood cells is two to three times longer than that of worker brood cells. The attractiveness of brood cells is related to the distance between the larva and the cell rim and the age of the larva. The moment of invasion of the mite into a brood cell is not related to the duration of its stay on adult bees. The fraction of the phoretic mites that invade brood cells is determined by the ratio of the number of suitable brood cells and the size of the colony. The distribution of mites over drone and worker brood in a colony is determined by the specific rates of invasion and the number of both brood types. Knowledge of mite invasion behaviour has led to effective biotechnical control methods and increased insight in the mite's population dynamics.<h4>Control of Varroa mites by combining trapping mites in honey bee worker brood with formic acid treatment of the capped brood outside the colony: Putting knowledge on brood cell invasion into practice (Chapter 2)</h4>Biotechnical <em>Varroa</em> control methods are based on the principle that mites inside brood cells are trapped and then removed from the bee colony. Initially, methods were studied in which worker brood was used for trapping. Trapped mites were killed with a formic acid treatment that left the worker brood unharmed. The observed percentage of mites trapped and killed by formic acid treatment was 87% and 89% in two experiments which matched predictions based on knowledge on brood cell invasion. Hence, knowledge on the mites' behaviour with respect to brood cell invasion proved to be a useful tool for designing and improving trap-comb methods for <em>Varroa</em> control.<h4>Effective biotechnical control of Varroa mites: Applying knowledge on brood cell invasion to trap mites in drone brood (Chapter 3)</h4>Trapping mites in brood cells is most efficient when drone brood is used while the colonies are otherwise broodless. In theory, one trap-comb using drone brood is enough to achieve effective control. I designed and tested two methods using trap-combs with drone brood. To reduce labour intensity, application of trap-combs was integrated in swarm prevention techniques. In the first method, effectiveness of the control method varied considerably, from 67% to 96%. Effectiveness depended on the number of drone cells that had been available for mite trapping. The observed effectiveness in each separate colony could be predicted from the numbers of bees and brood cells, thereby showing the validity of our approach. In the second method, we adjusted the method to improve production of drone brood on the trap-combs, because this appeared to be crucial for trapping efficiency. The observed effectiveness of 93.4 % demonstrates that trap-combs with drone brood can effectively trap mites, thereby offering a non-chemical method of <em>Varroa</em> control.<h4>Model evaluation of methods for Varroa mite control based on trapping in honey bee brood (Chapter 4)</h4>The trap-comb model that was used to predict mite-trapping effectiveness in our experiments was used to estimate and compare effectiveness of different trap-comb methods described by several authors. Predictions of the model showed that for effective control by trapping with worker brood is labour intensive because a large amount of brood is needed to trap a sufficient number of mites. An extra input of labour is the demand for treatment of the capped worker brood to selectively kill the mites, because beekeepers want to save the brood. The model predicted that trapping with drone brood demands much less brood cells for effective mite control. Labour intensity is less compared to trap-combs with worker brood. This is because drone brood with trapped mites is usually destroyed instead of saved and preparation of trap-combs with drone brood can be integrated into swarm-prevention-techniques.<h4>Population modelling of Varroa mites (Chapter 5)</h4>To understand population dynamics of the mite, Fries et al. (1994) incorporated knowledge on <em>Varroa</em> mite-honey bee interactions into a mite population model. I updated and extended this model by incorporating more recent data, in particular on mite invasion from bees into brood cells. This allowed predictions of invasion into and emergence from brood cells, and hence the distribution of mites over bees and brood. As mite control treatments usually only affect mites either in brood cells or on adult bees, the model can be used to evaluate their effectiveness and timing. Mite population growth proved to be especially sensitive to the length of the brood period, the number of drone cells and reproductive success in the brood cells.<h4>Natural selection of Varroa explains the different reproductive strategies in colonies of Apis cerana and Apis mellifera (Chapter 6)</h4>In colonies of European A. mellifera, Varroa reproduces both in drone and in worker brood. In colonies of its original Asian host, A. cerana, the mites invade both drone and worker brood cells, but reproduce only in drone cells. Absence of reproduction in worker cells is probably crucial for the tolerance of A. cerana towards Varroa because it means that the mite population can only grow during periods of drone rearing. To test whether the absence of mite reproduction in worker brood of A. cerana is due to a trait of the mites or of the honey bee species, mites from bees in A. mellifera colonies were introduced into A. cerana worker brood cells and vice versa. Approximately 80% of the mites originating from A. mellifera reproduced in worker cells of both A. mellifera and A. cerana. Conversely, only 10% of the mites originating from A. cerana colonies reproduced in worker cells of A. cerana and A. mellifera. Hence, absence of reproduction in worker cells is due to a trait of the mites. Additional experiments showed that A. cerana removed 84% of the worker brood that was artificially infested with mites from A. mellifera colonies. Brood removal started 2 days after artificial infestation, which suggests that the bees responded to behaviour of the mites. Because removal behaviour of the bees will have a large impact on the mite's fitness, it probably plays an important role in selection for differential reproductive strategies. These findings have large implications for selection programmes to breed less-susceptible bee strains. If differences in mites (i.e. whether they reproduce in worker brood or not) are mite-specific, we should not only look for mites not reproducing as such, but for colonies in which mites are selected for not reproducing in worker cells. Hence, in selection programmes reproductive success of mites that reproduce in both drone and worker cells should be compared to the reproductive success of mites that reproduce exclusively in drone cells.<h4>Reproductive success of Varroa mites in honey bee brood with differential development times (Chapter 7)</h4>Reproduction of <em>Varroa mites</em> has been extensively studied and many aspects of its life history such as number of eggs laid, timing of egg laying, and mortality of immature mites, are well known. However, estimates of the actual reproductive success after one brood cycle, i.e. how many mites can be found alive on the bees after emergence of an infested cell, are still fairly theoretical. Because this parameter is crucial for understanding population growth of the mites, several methods were used to measure the actual reproductive success. To evaluate how development time of the capped brood stage may affect population growth of the mites, measurements were done in bee strains with different development times of worker brood. In brood with a relatively short developmental time, reproductive success of mites was lower. Increased developmental time resulted in higher egg production and lower mortality of offspring before or shortly after emergence of the mites from the brood cell. The results show that the number of mites emerging alive from worker cells with relatively short development times, may become lower than the initial number that invaded the cells. This results in a decline of the mite population if only worker cells are available. In addition, the low reproductive success in worker brood with a short development time, explains that the phenomenon of mites not reproducing in worker cells, as found in <em>A. cerana</em> and in several <em>A. mellifera</em> races, evolves if these mites survive to reproduce in drone brood the next brood cycle.<h4>Attractiveness of brood cells of different honey bee races to Varroa mites (Chapter 8)</h4>Reproduction of the <em>Varroa</em> mite only occurs inside capped brood cells of honey bees. Therefore, invasion into brood cells is crucial for the mite's reproduction and the rate of invasion will affect the growth of the mite population. I investigated the invasion response of the mites to drone or worker larvae of different honey bee races, because selection for less attractive brood may help <em>Varroa</em> control. The observed differences in invasion response of <em>Varroa</em> mites to worker brood of the tested colonies were not statistically significant. The results suggest that not the racial origin of the worker brood, but the distance between the larva and the cell rim affects the invasion response of the <em>Varroa</em> mites to worker brood cells. Because measuring the distance between the larva and the cell rim in drone brood cells is inaccurate due to curved cell caps of neighbouring cells, the results for drone brood cells are difficult to interpret. Possibilities to obtain less attractive brood via selection or comb manipulation are discussed.<h3>Epilogue</h3><h4>Towards a future in which beekeeping does not depend on the use of acaricides for effective control of Varroa</h4>Considering the conflict between the use of synthetic acaricides and the status of honey bee products as natural products and the spreading resistance of <em>Varroa</em> to these acaricides, there is a clear need for alternative ways of <em>Varroa</em> control. Our research on biotechnical control methods and susceptibility of honey bees to <em>Varroa</em> contributes to sustainable <em>Varroa</em> control. Knowledge on invasion behaviour of mites into brood cells proved to be useful to understand the possibilities and limitations for improvement of biotechnical control methods. Using drone brood on trap-combs, an effective biotechnical control method has become available providing a non-chemical way of controlling the mite population. Integration of knowledge on invasion behaviour into a population model of the <em>Varroa</em> mite allows us to gain more insight in the mite's population dynamics and evaluate traits of honey bees that via selection may decrease susceptibility of honey bee colonies. Selection for honey bee traits that reduce reproductive success in worker brood in <em>A. mellifera</em> may lead to selection of mites towards the situation we know from the original host-parasite relationship were mites only reproduce in drone brood. The duration of the capped brood stage seems a good candidate because selection for a short development time will reduce reproductive success of the mites. Attractiveness of brood cells is a less suitable trait because differences in attractiveness of brood of different race were not detected. Although less susceptible honey bees are not available yet, selectable traits have been identified that may reduce the effect of <em>Varroa</em> infestation on honey bee colonies. Nowadays, beekeeping is not dependent on the use of synthetic acaricides to control the <em>Varroa</em> mite. Next to trap-comb methods, much research has been successfully directed towards <em>Varroa</em> control using organic acids and essential oils (Imdorf, 1999). Reducing susceptibility of honey bees together with effective control by means of biotechnical and other 'organic' control methods provides a perspective for beekeeping that does not rely on synthetic acaricides to kill <em>Varroa</em> mites.<h3>Acknowledgements</h3>I thank M. Beekman, WJ Boot, JC van Lenteren and M.W. Sabelis for their valuable comments on the manuscript.<h3>References</h3><dl><dt>Anderson, DL & Trueman, JWH (2000).</dt><dd><em>Varroa jacobsoni</em> (Acari: Varroidea) is more than one species. <em>Experimental and Applied Acarology</em> 24: 165-189.</dd><dt>Ball, BV (1994).</dt><dd>Host-Parasite-Pathogen interactions. <em>In</em> Matheson, A (editor) <em>New perspectives on Varroa</em> . IBRA, Cardiff, UK, pp 5-11.</dd><dt>Beetsma, J, de Vries, R, Emami Yeganeh, B, Emami Tabrizi, M & Bandpay, V (1989).</dt><dd>Effects of <em>Varroa jacobsoni</em> Oud.on colony development, workerbee weight and longevity and brood mortality. <em>In</em> Cavalloro, R (editor) <em>Present status of Varroatosis in Europe and progress in the Varroa mite control: proceedings of a meeting of the EC expert's group,</em> Udine, Italy, 28-30 November 1988, Commission of the European Communities, Luxembourg, pp 163-170.</dd><dt>Boot, WJ (1995).</dt><dd>Invasion of <em>Varroa</em> mites into honey bee brood cells. Thesis Wageningen University.</dd><dt>Boot, WJ, Tan, NQ, Dien, PC, Huan, LV, Dung, NV, Long, LT, & Beetsma, J (1997).</dt><dd>Reproductive success of <em>Varroa jacobsoni</em> in brood of its original host, <em>Apis cerana</em> , in comparison to that of its new host, <em>A</em> . <em>mellifera</em> (Hymenoptera: Apidae). <em>Bulletin of Entomological Research</em> 87: 119-126.</dd><dt>Büchler, R (1994).</dt><dd><em>Varroa</em> tolerance in honey bees -occurrence, characters and breeding. <em>Bee World</em> 75: 54-70.</dd><dt>Camazine, S (1986).</dt><dd>Differential reproduction of the mite, <em>Varroa jacobsoni</em> (Mesostigmata: Varroidae), on Africanized and European honey bees (Hymenoptera: Apidae). <em>Annals of the Entomological Society of America</em> 79: 801-803.</dd><dt>De Greef, M, Wael, L de & Laere, O van (1994).</dt><dd>The determination of the fluvalinate residues in the Belgian honey and beeswax. <em>Apiacta</em> 29: 83-87.</dd><dt>De Jong, D, Morse, RA & Eickwort, GC (1982).</dt><dd>Mite pests of honey bees, <em>Annual Review of Entomology</em> 27: 229-252.</dd><dt>Fries, I, Camazine, S & Sneyd, J (1994).</dt><dd>Population dynamics of <em>Varroa jacobsoni</em> : a model and a review. <em>Bee World</em> 75: 5-28.</dd><dt>Ifantidis, MD & Rosenkranz, P (1988).</dt><dd>Reproduktion der Bienemilbe <em>Varroa jacobsoni</em> (Acarina: Varroidae). <em>Entomologia Generalis</em> 14:111-112.</dd><dt>Imdorf, A, Charrière, JD & Fluri, P (1999).</dt><dd>Alternative strategies for the control of acaricide resistant <em>Varroa</em> mites. <em>Proceedings of the 36 ^th Apimondia congress</em> , Vancouver, Canada. pp 143.</dd><dt>Koeniger, N & Fuchs, S (1988).</dt><dd>Control of <em>Varroa jacobsoni</em> : current status and developments. <em>In</em> Needham, G R, et al. (editors) <em>Africanized honeybees and bee mites</em> . Ellis Horwood Limited, Chichester, pp 360-369.</dd><dt>Matheson, A (1993).</dt><dd>World bee health report. <em>Bee World</em> 74: 176-212.</dd><dt>Moritz, R (1994).</dt><dd>Selection for Varroatosis resistance in honeybees. <em>Parasitology Today</em> 10: 236-238.</dd><dt>Oudemans, AC (1904).</dt><dd>On a new genus and species of parasitic acari. <em>Notes from the Leyden Museum</em> 24: 216-222.</dd><dt>Ritter, W (1990).</dt><dd>Drug therapy development of <em>Varroa jacobsoni</em> in the federal republic of Germany. <em>Apiacta</em> 25: 102-106.</dd><dt>Ritter, W (1993).</dt><dd>New results of the development of tolerance to <em>Varroa jacobsoni</em> in bee colonies in Tunesia. <em>In</em> Connor, LJ et al. <em></em> (editors), <em>Asian Apiculture</em> . Wicwas Press, Cheshire, CT, USA, pp. 457-459.</dd><dt>Rosenkranz, P & Engels, W (1994).</dt><dd>Infertility of <em>Varroa jacobsoni</em> females after invasion into <em>Apis mellifera</em> worker brood as a tolerance factor against varroatosis. <em>Apidologie</em> 25:402-411.</dd><dt>Woyke, J (1989).</dt><dd>Breeding of honey bees resistant to <em>Varroa jacobsoni</em> . <em>American Bee Journal</em> 129: 21-23.</dd></dl>