The IACR-Rothamsted Varroa Project. by Norman L Carreck (norman.carreck@bbsrc.ac.uk) In the late 1970s, soon after Varroa jacobsoni was first discovered in honey bee colonies in central Europe, the mite was widely reported to be a devastating pest of Apis mellifera. However, observations from several countries suggested that the effects of mite infestation were variable. Some colonies seemed able to support very large mite populations without suffering apparent harm, whilst others collapsed and died out with fewer mites present. Since at this time the harmful effects of Varroa were attributed solely to the direct feeding action of the mite, the fact that damage was not always directly related to the level of infestation was difficult to explain. Long term studies at IACR-Rothamsted had shown that apparently healthy bee colonies invariably contained individuals with inapparent infections of a number of pathogens, especially viruses, and that there were associations between some virus infections and other parasites of bees1. For example, three viruses almost invariably multiply only in those individual bees which are also infected with Nosema apis. Two of these have been shown to add to the damaging effects of Nosema, and their presence or absence may account for the considerable variations in harm which have been attributed to the disease (2). It was therefore thought that differences in the incidence of secondary infections might explain some of the variability in the effects of Varroa. In the early 1980s, whilst working in Germany, Brenda Ball from IACR-Rothamsted analysed samples of dead bees collected from bee colonies with different degrees of Varroa infestation, for the presence of a number of known pathogens. The prevalence of some pathogens such as black queen cell virus (BQCV) and bee viruses X and Y (BVX, BVY) was found to be similar to that in mite-free colonies at IACR-Rothamsted. In contrast, chronic paralysis virus (CPV), the cause of the well known condition "bee paralysis", was much more abundant in moderately infested colonies in Germany, when compared to uninfested colonies in Britain. More surprisingly, dead bees from colonies heavily infested with mites were found to have very high levels of acute paralysis virus (APV) (3). This previously esoteric virus had been isolated in the 1970s during work on bee paralysis, but whilst commonly found as an inapparent infection in colonies, and demonstrated to be harmful when injected into bees in the laboratory, had never previously been found to cause mortality in nature. Studies showed that mites feeding on infected but symptomless adult bees in some way activated the virus to multiply, causing death. Furthermore, other work demonstrated the ability of the mite to act as a virus vector, transferring infection between adult bees, and to developing honey bee pupae. APV has since been found to be responsible for adult bee losses in Varroa infested colonies in Russia, USA, Yugoslavia, France and Greece. Later work led to the isolation of a previously unknown virus, deformed wing virus (DWV), from deformed newly emerged bees from colonies infested with mites, and demonstrated that the mite could act as a vector for a number of unrelated viruses (2,6). Following the discovery of Varroa in this country in the spring of 1992, we began studies of pathogen incidence in infested British colonies. In August and September 1992, in association with Keele University, samples of dead and live bees and hive debris were collected over a 28 day period from 44 colonies, some infested, some not, in the South Devon area. These were then tested for the presence of a range of viruses. Cloudy wing virus (CWV) and DWV were detected in a small number of dead bee samples, but APV was not, although there were indications that elevated levels of APV were present in live bees from some infested colonies. In the following spring a much larger sampling programme was begun. This project is funded by MAFF, and compliments the work being done by Dr Stephen Martin of the CSL National Bee Unit on the reproductive biology and population dynamics of the mite. The IACR-Rothamsted project aims to determine the causes of mortality in infested colonies, and the role of the mite in the epidemiology of honey bee pathogens. Between March 1993 and April 1994, samples of dead bees and hive debris were collected monthly from approximately 30 colonies at apiaries in South Devon. Live bees and diseased brood samples were also collected as appropriate. Samples were then tested for the known range of adult bee and brood diseases, including 11 viruses. These study colonies were managed by the beekeepers using their normal methods, and fell into three groups: firstly, "uninfested" colonies, either brought from IACR-Rothamsted, or located at sites, where Varroa had not been found in 1992; secondly infested colonies which had not been treated; and finally infested colonies which had been treated with an acaricide either in autumn 1992 or spring 1993. By the early summer the "uninfested" colonies were found to contain small numbers of mites, but all survived to April 1994 as apparently healthy colonies. The infested untreated colonies contained very large numbers of mites by the late summer, and with two exceptions were all dead by April 1994, although some losses were due to queenlessness and not necessarily due to Varroa infestation. The majority of the infested treated colonies survived to April 1994, although again some were lost through queenlessness. A few of these colonies, despite treatment, contained large numbers of mites by the late summer. The dead bees from all groups initially showed little incidence of virus infection, and no bees were found to be infected with APV or DWV. In late summer however, samples fromcolonies in one apiary, which had been treated with fluvalinate in August 1992, and which had unexpectedly large mite infestation, were found to contain large amounts of slow paralysis virus (SPV). This is another previously obscure virus, first isolated in the 1970s, which whilst shown in the laboratory to be fatal when injected into bees, had never before been found to cause mortality in nature. By January 1994, the two study colonies, and the remainder of the colonies in the apiary were all dead. SPV infection was later identified as a cause of adult bee and brood mortality in six of the study colonies in four apiaries. The first year of the project therefore confirmed the variable nature of the effects of Varroa jacobsoni infestation, which are almost certainly more variable in this country than on mainland Europe, possibly due to our more unpredictable weather. The discovery of SPV as a cause of mortality in infested colonies in Devon has further emphasised the ability of the mite to activate a normally latent virus to multiply to lethal levels, making prediction of its harmful effects difficult. During 1994, our research concentrated on six infested untreated colonies in Hertfordshire, which were intensively studied. Every three weeks, dead bees were collected from entrance traps, dead mites were collected from the floor debris, and total adult bee and brood populations were estimated. Approximately 1% of the live adult bees, and 1% of the sealed brood were also collected to measure their degree of infestation. This enabled the total mite population to be estimated, and to be compared with the bee and brood populations. Samples of live and dead bees, infested and uninfested brood and mites were then saved and analysed for the presence of a range of pathogens, especially viruses. This information showed that from a small number of mites present in the early spring, populations can increase rapidly, during the period of maximum bee population growth. Mitepopulations peaked in late summer, and then rapidly declined, as has been observed in other colonies in the UK (9). This decline occurs at the time when honey bee brood rearing is drastically reduced, and when there is excessive competition amongst mites for brood cells in which to reproduce (7). We, have observed that a large proportion of dead mites at this time of year are pale in colour, indicating immaturity and mortality before reproducing. This phenomenon has also been noted by other researchers (10), and requires further study. Our studies have suggested that the rate of increase of the bee population is a far more important factor affecting mite population growth than the number of mites initially present. In a weak colony, with erratic brood laying, or under poor weather conditions, mite numbers may increase very little over a season. In contrast, in a large colony with a young queen, and under good weather conditions, mite populations can increase rapidly. What appears to have happened in 1995 is that the exceptionally good summer favoured both the bee and the mite, allowing mite populations to increase rapidly from low levels to very high levels, prompting the appearance of virus infections and colony losses. Virus analysis of the 1994 samples showed that SPV was a primary cause of adult bee and brood mortality, as in the previous year. Some dead brood samples were also found to contain much cloudy wing virus (CWV), previously known only as a disease of adult bees, and sacbrood virus (SBV) was also commonly found early in the season in infested colonies. SPV could not be detected by sensitive serological tests in live bees or brood any earlier than it was detected in dead bees. In contrast, SPV was detectable in mites two months before it could be detected in the bees, implying that early prediction of virus infections in colonies may be possible, which would allow timely treatment of the mites to prevent colony loss. In 1995, our work concentrated on laboratory studies in order to quantify the directdamaging effects of mite feeding on pupae and adult bees. Studies elsewhere have correlated increasing mite infestation with increasing mortality of pupae (8), but these did not take account of the possible presence of pathogens such as viruses. Our studies, using bees and mites in the absence of viruses, have shown that the direct effects of mite feeding on both adult bees and brood appear to be minimal. Bees emerging from infested cells initially have a lower body weight (20 % less with five mites per cell) than those from uninfested cells, but these differences soon disappear, and we have found no significant differences between the longevity of bees emerging from infested and uninfested cells in cage experiments. The variability in the damaging effect of mite infestations, and in the mortalities caused by honey bee virus infection may be due to differences in virus incidence in infested colonies, but equally may be dependent on innate differences in susceptibility to virus infection between honey bee populations. Investigation of the response by the host to parasitism and of the factors inducing virus replication may identify differences between honey bee populations which could contribute to breeding or selection programmes aimed at increasing the tolerance of honey bees to Varroa. Our other studies also have practical applications. More precise timing and targeting of acaricide treatments will reduce losses due to secondary infections and minimise chemical inputs into colonies. The mites themselves are probably sensitive indicators of virus presence in colonies, and the development of a predictive test could help to identify those colonies most likely to suffer damage. In addition, further work on the effect of the timing of acaricide treatment on virus epidemiology would define more accurately when action was required. As with most biological systems, the process of infestation by Varroa is affected by many conflicting factors. The growth of the mite population depends on the population of adult beesand brood, which in turn is dependant on weather conditions, forage availability, the degree of competition from other bees, the age of the queen, and presence of disease. Secondary virus infections add further variability to the system. We can, however, be certain that the final death of a colony is unlikely to be due to the effects of Varroa alone, or due to the effects of a single pathogen. Only by investigating and evaluating the factors involved can we hope to understand the process of infestation, which is essential to devise effective control strategies for the future. References 1: Bailey L, Ball B V, Perry J N (1981) The prevalence of viruses of honey bees in Britain. Annals of Applied Biology 97 109-118. 2: Bailey L, Ball B V (1991) Honey bee pathology (2nd ed.). Academic Press. 193pp. 3: Ball B V (1985) Acute Paralysis Virus isolates from honey bee colonies infected with Varroa jacobsoni. Journal of Apicultural Research 24 115-119. 4: Ball, B.V. (1993) The damaging effects of Varroa jacobsoni infestation. in Living with Varroa. ed. A. Matheson. IBRA; Cardiff. pp 9-16. 5: Ball, B.V. (1994) Host-parasite-pathogen interactions. in New perspectives on Varroa. ed A. Matheson. IBRA; Cardiff. pp 5-11. 6: Ball B V, Allen M F (1988) The prevalence of pathogens in honey bee colonies infested with the parasitic mite Varroa jacobsoni. Annals of Applied Biology 113 237-244. 7: Fuchs, S., Langenbach, K. (1989) Multiple infestation of Apis mellifera L. brood cells and reproduction in Varroa jacobsoni Oud. Apidologie 20, 257-266. 8: Kovac, H, Crailsheim, K. (1988) Lifespan of Apis mellifera carnica Pollm. infested by Varroa jacobsoni Oud. in relation to season and extent of infestation. Journal of Apicultural Research 27, 230-238. 9: Martin, S.J., Hogarth, A.N. (1994) Varroa research carried out by the NBU in the UK. in New Perspectives on Varroa. ed. A.Matheson. IBRA; Cardiff. pp 38-40. 10: Rademacher, E. (1985) [Is a prediction of infestation by Varroa jacobsoni possible based on its natural death rate?] (in German) Apidologie 16, 365-406. HTML conversion by Colin H Denholm [Image] © IACR-Rothamsted 1997 Research Areas | Projects | Staff | Students | Publications | Gallery | WWW Resources | Varroa WWW Hub | Useful Links