The bee, the mite, and the virus

Bee mite

Editor's Introduction

Global honey bee viral landscape altered by a parasitic mite

annotated by
Oya Cingöz

The large-scale death of honey bee colonies worldwide has caused significant financial and ecological losses over the past decade. Increased prevalence of an invasive parasitic mite, the aptly named Varroa destructor, has been linked to honey bee colony deaths. It turns out that mite infestation is not the whole story, however. Varroa acts as a viral amplifier by augmenting the viral load present in the bees and the colony as a whole, causing large-scale colony collapse. This study examines the association between Varroa infestation in bee colonies in Hawaii and the rise of a single dominant strain of Deformed Wing Virus (DWV).  The co-occurrence of this mite and the selected virus strain may provide an explanation to the mysterious deaths of honey bee colonies worldwide.

Paper Details

Original title
Global honey bee viral landscape altered by a parasitic mite
Stephen J. Martin Ethel M. Villalobos et al.
Original publication date
Vol. 336 no. 6086 pp. 1304-1306
Issue name


Emerging diseases are among the greatest threats to honey bees. Unfortunately, where and when an emerging disease will appear are almost impossible to predict. The arrival of the parasitic Varroa mite into the Hawaiian honey bee population allowed us to investigate changes in the prevalence, load, and strain diversity of honey bee viruses. The mite increased the prevalence of a single viral species, deformed wing virus (DWV), from ~10 to 100% within honey bee populations, which was accompanied by a millionfold increase in viral titer and a massive reduction in DWV diversity, leading to the predominance of a single DWV strain. Therefore, the global spread of Varroa has selected DWV variants that have emerged to allow it to become one of the most widely distributed and contagious insect viruses on the planet.


The emergence of infectious diseases is driven largely by socioeconomic, environmental, and ecological factors (1), and these diseases have significant effects on biodiversity, agricultural biosecurity, global economies, and human health (23). The honey bee is one of the most economically important insects, providing crop pollination services and valuable hive products (4). During the past 50 years, the global spread of the ectoparasitic mite Varroa destructor has resulted in the death of millions of honey bee (Apis mellifera) colonies (5). There is general consensus that the mites’ association with a range of honey bee RNA viruses is a contributing factor in the global collapse of honey bee colonies (510), because the spread of mites has facilitated the spread of viruses (11, 12) by acting as a viral reservoir and incubator (13). In addition, the mites’ feeding behavior allows virus to be transmitted directly into the bees’ hemolymph, thus bypassing conventional, established oral and sexual routes of transmission. In particular, deformed wing virus (DWV) has been associated with the collapse of Varroa-infested honey bee colonies (5, 8, 14–16), because it is ubiquitous in areas where Varroa is well established (691718). The rapid global spread of Varroa means that very little is known about the natural prevalence, viral load, and strain diversity of honey bee viruses before the Varroa invasion (15). Such data are important, because most honey bee viral infections were considered harmless before the spread of Varroa (9). Large-scale loss of honey bee colonies has been associated with viruses vectored by Varroa (5). The recent arrival and subsequent spread of Varroa across parts of the Hawaiian archipelago has provided an opportunity to study the initial phase of the evolution of the honey bee–Varroa–DWV association. So far, colony collapse disorder (CCD) (6) has not been reported in Hawaii (19), but all of the associated pests and pathogens are present.

European honey bees (Apis mellifera L.) were first introduced to Hawaii from California in 1857. They were largely managed, but feral populations were soon established on every major island in the archipelago (20). Hawaii remained Varroa-free until August 2007, when the mite was discovered throughout Oahu Island. A subsequent survey by S. Nikaido and E. Villalobos during 2007–2008 recorded the collapse of 274 of 419 untreated colonies belonging to beekeepers. The disappearance of feral colonies from urban areas on Oahu was also noticed by beekeepers and pest control officers. Despite quarantine measures, the mite spread to Hilo on the Big Island in January 2009, where it survived an eradication attempt and by November 2009 had spread throughout the southern region of the island (Fig. 1). By November 2010, Varroa occurred throughout the Big Island. However, the islands of Kauai and Maui remained mite-free, and no unusual colony losses or disease problems have been reported there (19). The aim of this study was to investigate the influence that Varroa has in the spread of honey bee viruses during the initial phase of establishment. The spread of Varroa is normally from point introductions characteristic of pest species, so the arrival and spread of the mite across Hawaii are typical for this species.


Fig. 1.  (A) The four main Hawaiian Islands, showing the distribution ofVarroa during 2009. Green and brown indicate Varroa-free andVarroa infested areas respectively. Dots indicate the location of each study apiary. By November 2010, Varroa was present throughout the Big Island. The co-occurrence of the Varroa mite (B) and DWV can result in overt symptoms of (C) deformed wings in honey bees, although many nondeformed bees also carry high DWV loads.


General information about the sample collection sites in Hawaii, the presence/absence of the mite, and photographs of the host (honey bee; Apis mellifera), the mite (Varroa destructor), and the virus (Deformed Wing Virus; DWV).


Geographical distribution of the Varroa mite among the honey bee colonies in the four main Hawaiian islands in 2009.

Green: Varroa-free

Red: Varroa-infested.

The mite and the virus 

Electron micrographs (microscopy photographs) of the Varroa mite (left) and the Deformed Wing Virus (DWV; right).

More information on electron microscopy can be found here

The University of Hawaii Electron Microscopy website offers interesting images of insects and other small organisms


A photograph showing deformed wings in honey bees caused by the co-occurrence of DWV and Varroa

In 2009, our study of 293 honey bee colonies, from 35 apiaries on the four main Hawaiian Islands (Fig. 1), revealed that the exposure to Varroa had a significant effect on the prevalence, viral load, and strain diversity of DWV (Fig. 2). In contrast, neither the prevalence nor the viral load of any of the other four viruses investigated [Kashmir bee virus (KBV), slow paralysis virus (SPV), acute bee paralysis virus (ABPV), or Israeli acute paralysis virus (IAPV)] was affected by the presence of Varroa (fig. S1).


Fig. 2.  Viral load, prevalence, and genetic diversity of DWV across the four main Hawaiian Islands that have been exposed to Varroa for different periods of time. 1.E + 04 = 1 × 104, etc. The asterisk indicates a Varroa-free feral colony that died. Red indicates the proportion of positive colonies (supported by two or more positive RT-PCR tests) in the DWV prevalence pie charts, with the total number of colonies sampled from each population shown beneath. Strain diversity is based on HRM analysis of three randomly selected colonies from each population (figs. S2 and S5) and is supported by the rarefaction curves (fig. S4).


The graphs depict the viral load, prevalence, and strain diversity of DWV in bee colonies from the four main islands of Hawaii.

These colonies have been exposed to the Varroa mite for different lengths of time, and they show significant differences in the DWV viral load, prevalence, and strain diversity, based on the time that has passed since Varroa infestation.

Viral load

Viral load is the amount of virus present in the hemolymph of a bee.

The viral load can be determined by quantifying the amount of viral RNA sequences present in each bee by RT-PCR.

The island of Oahu, which has been exposed to Varroa longest (more than 3 years), has the highest average viral load per bee.

The second highest viral load is seen in colonies from the Big Island in 2010, exposed to  for less than 2 years.

Colonies in other regions that have been exposed to Varroa for less than 1 year, or not at all, have similarly low levels of viral RNA per bee.

This graph therefore shows that the length of Varroa exposure is associated with increased viral loads.


The prevalence of DWV in different bee colonies is indicated by the red color. Exposure to Varroa increases the prevalence of DWV significantly. In Oahu, where colonies have been exposed to Varroa for more than 3 years, 100% of the colonies also carry DWV.

Strain diversity

Each color in the chart represents a different strain of DWV; the more colorful the chart, the greater the viral diversity.

Bee colonies that have been exposed to Varroa for less than a year (the Big Island in 2010), or not at all (Kaui, Maui, and the Big Island in 2009) have the greatest DWV strain diversity.

Colonies that have been infested for 2 years or more have significantly lower virus diversity, e.g., Oahu colonies only carry only one dominant DWV strain.


Exposure to Varroa causes the prevalence of DWV to increase from 6% to 13% to 96% to 100% among bee colonies.

This increase in prevalence is accompanied by about a million-fold increase in the viral load per bee.

Despite the increased amount of virus in the individuals and the population, the strain diversity decreases significantly following exposure to Varroa

In Varroa-free areas, DWV was detected in 6 to 13% of colonies, but it increased to 75 to 100% where Varroa had been established (Fig. 2). Increased DWV prevalence was accompanied by a millionfold difference in viral load between Varroa-free areas (<1000 DWV copies per bee) and Varroa-infested areas (>1,000,000,000 DWV copies per bee) (Fig. 2), although there was a time lag associated with changes in strain diversity (that is, between 2009 and 2010 on the Big Island) (Fig. 2 and fig. S2). High-resolution melting (HRM) analysis of DWV–reverse transcription polymerase chain reaction (RT-PCR) products showed that in 2009, 20 colonies from five apiaries, each maintained by independent bee farmers on Oahu, were primarily dominated by a single genotype cluster (Fig. 2 and fig. S2), and sequencing showed that this sequence was identical to DWV sequences previously detected in the United Kingdom, Italy, Denmark, Spain, and France (fig. S3). In 2009, the HRM profiles for Kauai, Maui, and Big Island samples exhibited multiple peaks, indicating the presence of a range of DWV variant sequences (Fig. 2 and fig. S2). Rarefaction analysis of DWV diversity on each island confirmed these findings, with the cumulative number of strains reaching saturation in areas where Varroa had been established. In recently invaded or Varroa-free regions, the rarefaction curves did not approach saturation, which is typical of highly diverse systems (fig. S4). One year later, Varroa had spread across the Big Island, and a follow-up study of 38 colonies from six apiaries showed the same pattern as previously seen on Oahu: an increase in viral load and a decrease in variant diversity (Fig. 2 and figs. S2 and S4). After 1 year of effective Varroa control on Oahu, data from 11 colonies in one apiary in 2010 indicated that the same DWV strain remained dominant (fig. S2), suggesting that Varroa-induced changes to the viral landscape are capable of persisting despite the Varroa populations being under control.

Using 40 clones, sequence analysis revealed 10 virus variants in single bee colonies from each of the four islands, with Kauai, Maui, and the Big Island each having a unique DWV variant (fig. S5). This indicated that a single colony from a Varroa-free area contained more viral diversity than that detected across Oahu or the Big Island (in 2010) after Varroa had become well established. Subsequent analysis of sequence data separated two distinguishable DWV variant groups (figs. S3 and S5): (i) the “classic” DWV sequence known from symptomatic and asymptomatic honey bees in both Varroa-free and Varroa-infested colonies; and (ii) a DWV sequence sharing approximately 18 nucleotide substitutions with the closely related Varroa destructor virus (VaDV-1) and only 82% sequence homology to the classic DWV sequence (fig. S3).

Varroa populations are largely controlled by the use of pesticides, but depending on the season, nearly all bee colonies are infected by DWV (9). Such observations are probably due to the fact that Varroa is never fully eradicated from infested colonies, and vertical transmission through males (drones) and queens exists (21).Varroa’s arrival at Hawaii has fundamentally altered the viral landscape in both managed and feral bee colonies. On Oahu, all six feral colonies tested had high levels of DWV (6.1 × 108 copies per bee), similar to that found in managed colonies (Fig. 2), whereas only one of nine feral colonies from the Varroa-free area of the Big Island carried DWV, and this was the only honey bee colony in a Varroa-free area with a high viral level (4.6 × 107 copies per bee) (Fig. 2). This colony had a similar melt curve to that produced by the Oahu cluster and subsequently died within a year. High DWV loads (>107 copies per bee) have also been associated with colony death in Varroa-free areas, indicating that naturally virulent variants can cause colony death, with or without signs of wing deformity, although rarely (15).

Variant group A virus is usually associated with symptomatic DWV in the presence of Varroa, although it has also been found in Varroa-free colonies at significantly lower levels, such as those from Kauai. Rather than resulting from recent recombination events, the B variant and putative DWV/VaDV-1 hybrids (22) may simply represent sequence variants that have always existed, and perhaps they highlight the extent of the natural genetic diversity within the collective DWV variants. Furthermore, we found that the Oahu variants had a greater similarity to DWV than to VaDV-1 at regions sequenced on both sides of the proposed recombination points (22) (fig. S6). Additionally, leader protein sequence data from Oahu variants clustered with DWV and not VaDV-1 sequences (fig. S7). These findings indicate that the increase in viral load on Oahu is not a result of the formation of DWV/VaDV-1 hybrids. The replicative form of DWV has been detected in mites (23), and this study indicates that the presence of Varroa over time is selecting for particular variants that may give them a competitive advantage. In Hawaii, the main Oahu strain was also detected in colonies from the Big Island, Maui, and Kauai during 2009 but at much lower frequencies (Fig. 2), supporting the hypothesis that Varroa facilitates the dominance of certain strains (23), which is strengthened by the loss of strain diversity between 2009 and 2010 as Varroa became established on the Big Island (Fig. 2 and fig. S2). Many factors are likely to influence the DWV variant population in different colonies, but the arrival of DWV variants that can replicate in the mite (13) means that these strains would rapidly increase in abundance. There have been no major introductions of honey bees into Hawaii, because strict importation regulations have been enacted since the widespread occurrence of Varroa mites. It seems likely that the now mite-associated European DWV variants were already present in honey bee populations before the arrival of the mites. Studies in the United Kingdom (14) and New Zealand (24) have found that DWV infections and colony collapse did not coincide with the arrival and establishment of Varroa, but there was with a 1- to 3-year time lag, which we also observed on Hawaii. This lag appears to be the time required for the selection of virus variants adapted to mite transmission.

Recent studies have found no correlation between the presence of Varroa and changes in host immune responses (102526), and the common occurrence of time lags between mite introduction and establishment suggests that the increase in DWV titer and reduction in variant diversity cannot be explained by Varroa-induced immunosuppression of honey bees (27). The apparent lack of association between ABPV, IAPV, and KBV and Varroa in this study may reflect the fact that the latter viruses require a longer lag period to become established in Varroa than does DWV, although the prevalence of these viruses varies greatly in Varroa-infected areas. Further work is required to elucidate the precise role that Varroa may have in influencing the prevalence of the range of viruses that infect bees and their role in colony collapse.

Complete viral genome sequencing and experimental infections of honey bees with different DWV strains are required for testing virulence and Varroa-associated honey bee colony losses as was seen on Oahu and the Big Island. The current Varroa-adapted DWV variants will continue to evolve, and investigations of virus strain differences may explain the different pathologies currently seen globally in honey bee colonies (7). Such variants may interact with other pests, pathogens, environmental factors, and regional beekeeping practices, resulting in recent large-scale losses of honey bee colonies (6). This study shows that the spread of Varroa in Hawaii has caused DWV, originally an insect virus of low prevalence, to emerge. This association may be responsible for the death of millions of colonies worldwide wherever Varroa and DWV co-occur.

Supplementary Materials

Materials and Methods
Table S1
Figs. S1 to S7
References (2837)

References and Notes

  1. R. Blaustein, P. T. J. Johnson, Conservation biology: When an infection turns lethal. Nature 465, 881 (2010).

  2. K. E. Jones et al., Global trends in emerging infectious diseases. Nature 451, 990 (2008).

  3. J. K. Waage, J. D. Mumford, Agricultural biosecurity. Philos. Trans. R. Soc. London Ser. B 363, 863 (2008).

  4. R. A. Morse, N. W. Calderone, The value of honey bees as pollinators of U.S. crops in 2000. Bee Culture 128, 1 (2000)

  5. S. J. Martin, The role of Varroa and viral pathogens in the collapse of honey bee colonies: A modelling approach. J. Appl. Ecol. 38, 1082 (2001).

  6. D. L. Cox-Foster et al., A metagenomic survey of microbes in honey bee colony collapse disorder. Science 318, 283 (2007).

  7. C. Highfield et al., Deformed wing virus implicated in overwintering honeybee colony losses. Appl. Environ. Microbiol. 75, 7212 (2009).

  8. N. L. Carreck, B. V. Ball, S. J. Martin, Honey bee colony collapse and changes in viral prevalence associated with Varroa destructor. J. Apic. Res. 49, 93 (2010).

  9. E. Genersch, M. Aubert, Emerging and re-emerging viruses of the honey bee (Apis mellifera L.). Vet. Res. 41, 54 (2010).

  10. R. M. Johnson, J. D. Evans, G. E. Robinson, M. R. Berenbaum, Changes in transcript abundance relating to colony collapse disorder in honey bees (Apis mellifera). Proc. Natl. Acad. Sci. U.S.A. 106, 14790 (2009).

  11. P. L. Bowen-Walker, S. J. Martin, A. Gunn, The transmission of deformed wing virus between honey bees (Apis mellifera L.) by the ecto-parasitic mite Varroa jacobsoni. Oud. J. Invertebr. Pathol. 73, 101 (1999).

  12. D. Sumpter, S. J. Martin, The dynamics of virus epidemics in Varroa-infested honey bee colonies. J. Anim. Ecol. 73, 51 (2004).

  13. S. Gisder, P. Aumeier, E. Genersch, Deformed wing virus: replication and viral load in mites (Varroa destructor). J. Gen. Virol. 90, 463 (2009).

  14. S. J. Martin, A. Hogarth, J. van Breda, J. Perrett, A scientific note on Varroa jacobsoni Oudemans and the collapse of Apis mellifera colonies in the United Kingdom. Apidologie (Celle) 29, 369 (1998).

  15. J. R. de Miranda, E. Genersch, Deformed wing virus. J. Invertebr. Pathol. 103 (suppl. 1), S48 (2010).

  16. E. Genersch et al., The German bee monitoring project: A long term study to understand periodically high winter losses of honey bee colonies. Apidologie (Celle) 41, 332 (2010).

  17. D. Tentcheva et al., Prevalence and seasonal variations of six bee viruses in Apis mellifera L. and Varroa destructor mite populations in France. Appl. Environ. Microbiol. 70, 7185 (2004).

  18. C. Baker, D. C. Schroeder, Occurrence and genetic analysis of picorna-like viruses infecting worker bees of Apis mellifera L. populations in Devon, South West England. J. Invertebr. Pathol. 98, 239 (2008).

  19. S. J. Martin, Trouble in paradise: Varroa spreads across Hawaii. Am. Bee J. 150, 381 (2010).

  20. K. M. Roddy, L. A. Arita-Tsutsumi, History of honey bees in the Hawaiian Islands. J. Hawaiian Pacific Agriculture 8, 59 (1997).

  21. Yue, M. Schröder, S. Gisder, E. Genersch, Vertical-transmission routes for deformed wing virus of honeybees (Apis mellifera). J. Gen. Virol. 88, 2329 (2007).

  22. J. Moore et al., Recombinants between Deformed wing virus and Varroa destructor virus-1 may prevail in Varroa destructor-infested honeybee colonies. J. Gen. Virol. 92, 156 (2011).

  23. Yue, E. Genersch, RT-PCR analysis of Deformed wing virus in honeybees (Apis mellifera) and mites (Varroa destructor). J. Gen. Virol. 86, 3419 (2005).

  24. J. H. Todd, J. R. de Miranda, B. V. Ball, Incidence and molecular characterization of viruses found in dying New Zealand honey bee (Apis mellifera) colonies infested with Varroa destructor. Apidologie (Celle) 38, 354 (2007).

  25. P. G. Gregory, J. D. Evans, T. Rinderer, L. de Guzman, Conditional immune-gene suppression of honeybees parasitized by Varroa mites. J. Insect Sci. 5, 7 (2005).

  26. M. Navajas et al., Differential gene expression of the honey bee Apis mellifera associated with Varroa destructor infection. BMC Genomics 9, 301 (2008).

  27. X. Yang, D. L. Cox-Foster, Impact of an ectoparasite on the immunity and pathology of an invertebrate: Evidence for host immunosuppression and viral amplification. Proc. Natl. Acad. Sci. U.S.A. 102, 7470 (2005).

  28. Ratti et al., Detection and relative quantitation of soil-borne cereal mosaic virus (SBCMV) and Polymyxa graminis in winter wheat using real-time PCR (TaqMan). J. Virol. Methods 122, 95 (2004).

  29. R. Siede, M. König, R. Büchler, K. Failing, H.-J. Thiel, A real-time PCR based survey on acute bee paralysis virus in German bee colonies. Apidologie (Celle) 39, 650 (2008).

  30. E. Genersch, Development of a rapid and sensitive RT-PCR method for the detection of deformed wing virus, a pathogen of the honeybee (Apis mellifera). Vet. J. 169, 121 (2005).

  31. O. Berényi et al., Phylogenetic analysis of deformed wing virus genotypes from diverse geographic origins indicates recent global distribution of the virus. Appl. Environ. Microbiol. 73, 3605 (2007).

  32. K. Tamura et al., MEGA5: Molecular evolutionary genetics analysis using maximum likelihood, evolutionary distance, and maximum parsimony methods. Mol. Biol. Evol. 28, 2731 (2011).

  33. N. Zioni, V. Soroker, N. Chejanovsky, Replication of Varroa destructor virus 1 (VDV-1) and a Varroa destructor virus 1-deformed wing virus recombinant (VDV-1-DWV) in the head of the honey bee. Virology 417, 106 (2011).

  34. P. Chantawannakul, L. Ward, N. Boonham, M. A. Brown, A scientific note on the detection of honeybee viruses using real-time PCR (TaqMan) in Varroa mites collected from a Thai honeybee (Apis mellifera) apiary. J. Invertebr. Pathol. 91, 69 (2006).

  35. R. Kajobe et al., First molecular detection of a viral pathogen in Ugandan honey bees. J. Invertebr. Pathol. 104, 153 (2010).

  36. J. R. de Miranda et al., Genetic characterization of slow bee paralysis virus of the honeybee (Apis mellifera L.). J. Gen. Virol. 91, 2524 (2010).

  37. M. Allen, B. V. Ball, The incidence and world distribution of honey bee viruses. Bee World 77, 141 (1996).

  38. Acknowledgments: We thank all the Hawaiian beekeepers that participated in this study and D. Jackson of Sheffield University and C. Godfray of Oxford University for comments. S.J.M. and L.B. were funded by a Natural Environment Research Council urgency grant (NE/H013164/1), an Organisation for Economic Co-operation and Development fellowship, and the C. B. Dennis Trust. G.E.B. and M.P. were funded by a grant from the Waterloo foundation with support from Defra and the Welsh Assembly Government. D.C.S. and A.C.H. were funded by the C. B. Dennis Trust. S.N. and E.M.V. were funded by the U.S. Department of Agriculture’s National Institute of Food and Agriculture Tropical and Subtropical Agricultural Research (TSTAR) Program (grant no. 2010-34135-21499) and by support from the Hawaii Department of Agriculture. The data are available in the DRYAD depository at