Implications of horizontal and vertical for honey bee Ingemar Fries, Scott Camazine

To cite this version:

Ingemar Fries, Scott Camazine. Implications of horizontal and vertical pathogen transmis- sion for honey bee epidemiology. Apidologie, Springer Verlag, 2001, 32 (3), pp.199-214. ￿10.1051/apido:2001122￿. ￿hal-00891679￿

HAL Id: hal-00891679 https://hal.archives-ouvertes.fr/hal-00891679 Submitted on 1 Jan 2001

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Apidologie 32 (2001) 199–214 199 © INRA/DIB-AGIB/EDP Sciences, 2001

Review article

Implications of horizontal and vertical pathogen transmission for honey bee epidemiology

Ingemar FRIESa*, Scott CAMAZINEb

a Department of Entomology, Swedish University of Agricultural Sciences, Box 7044, 750 07 Uppsala, Sweden b Department of Entomology, Penn State University, 539 ASI Building, University Park, PA 16802, USA

(Received 15 January 2001; accepted 8 March 2001)

Abstract – The degree to which a disease evolves to be virulent depends, in part, on whether the pathogen is transmitted horizontally or vertically. Eusocial insect colonies present a special case since the fitness of the pathogen depends not only on the ability to infect and spread between individuals within a colony, but also on the ability to spread to new individuals in other colonies. In honey bees, intercolony transmission of occurs horizontally (by drifting or robbing) and vertically (through swarming). is likely the most important route of pathogen of new colonies. Theory predicts that this should generally select for benign -parasite relation- ships. Indeed, most honey bee diseases exhibit low . The only major exception is American foulbrood (AFB). In light of current ideas in evolutionary epidemiology, we discuss the implica- tions of horizontal and vertical pathogen transmission for virulence of AFB and other honey bee diseases. honeybee / epidemiology / pathogen transmission / / vertical transmission

1. INTRODUCTION another (see, for example Hurd and Lane, 1998 and references therein). Parasites In the coevolution of hosts and their par- evolve specialized mechanisms for trans- asites, both the parasite and the host may mission either between individuals within gain in fitness by manipulating the behavior, a single generation (horizontal transmis- physiology and life-history traits of one sion), between individuals between one

* Correspondence and reprints E-mail: [email protected] 200 I. Fries, S. Camazine generation and the next (vertical transmis- consequence of the parasite’s reproduction. sion), or both, and these differences in mode However, here is a trade off for the para- of transmission greatly affect host fitness site. If the parasite’s reproduction is too (Lipsitch et al., 1995a, b). high, the infection may be so virulent that the host dies before sufficient numbers of Here we discuss host-parasite relation- propagules can be transmitted to new hosts. ships and modes of pathogen transmission in On the other hand, if parasite reproduction honey bees. We restrict our discussion to a is too low, the parasite loses opportunities social insect, the honey bee, because of our for transmission. The observed differences interest in examining the implications of in parasite virulence in honey bees are more host-parasite interactions in the special sit- likely to be adaptions on the part of the par- uation presented by sociality. We focus on asites rather than counterdefenses on the honey bees because they are the most well- part of the host, since parasites generally studied social insect, and because of the have much shorter generation time and potential economic and applied implications higher mutation rates than their hosts. This of better understanding honey bee epi- enables the parasite to evolve faster than the demiology. host (Hafner et al., 1994). Nonetheless, it Honey bee colonies consist of three is often difficult to distinguish changes in castes; workers which number about parasite virulence from changes in host resis- 15000–50000, a few hundred drones and tance to parasites (Schmid-Hempel, 1998). one reproducing female, the queen. A honey A number of factors influence the evo- bee colony grows through the reproductive lution of virulence. The mode of disease efforts of a single queen. In this haplodiploid transmission is one such factor believed to system, fertilized eggs become non-repro- play a crucial role in molding pathogen ductive workers who help raise their sisters, virulence over evolutionary time (Lipsitch while unfertilized eggs become haploid male et al., 1996). The different modes of drones. Although the colony also may pathogen transmission can be divided into achieve some reproductive fitness through two categories: horizontal and vertical trans- its drones, the production of new queens by mission. Horizontal transmission refers to workers and the subsequent division into parasite transmission between individuals two or several colonies, is fundamental for of the same generation, while vertical trans- colony fitness (Moritz and Southwick, mission refers to that from parent to off- 1992). It is this reproductive swarming, spring (Canning, 1982). Compared with hor- which produces new colonies that makes izontal transmission, vertical parasite honey bees a special case when discussing transmission is expected to select for honey bee epidemiology and disease trans- decreased virulence. In this situation the mission. goals of the host and parasite are aligned. Successful vertical transmission of a parasite requires that the host reproduces effectively, 2. DISEASE VIRULENCE and thus the parasite reduces its own fitness AND PATHOGEN TRANSMISSION if it adversely affects the overall health and reproductive capacity of its host. This idea Virulence can be defined as the degree has been tested experimentally and con- to which parasitic infection decreases host firmed in a system using bacteria and bacte- survival and reproduction. Virulence is adap- riophages transmitted either vertically or tive for the parasite only insofar as it pro- horizontally (Bull et al., 1991). Observa- motes its own fitness, for example by pro- tions on fig wasps have also demonstrated a moting transmission of parasites to new close correlation between nematode viru- hosts; often virulence is an unavoidable lence and degree of horizontal transmission Honeybee epidemiology 201 relative to vertical transmission (Herre, (3) Host longevity. In short-lived hosts, 1993). Furthermore, experiments using the parasite must “race against the clock” 500 generations of a bacterial culture have to assure transmission to a new host. In this also demonstrated a genetic trade-off situation, there is little disadvantage to the between vertical and horizontal transmis- rapid development of large parasite popu- sion of plasmids, although increased viru- lations (high virulence). In contrast, even lence due to horizontal transmission could when the parasite is not virulent, a long- not be demonstrated in this experiment lived host provides many opportunities for (Turner et al., 1998). It should be noted that disease transmission (Lipsitch et al., 1996). model simulations indicate that where both (4) Population structure. If there is suffi- horizontal and vertical transmission occur, cient structure in the host population, one the expected virulence is not simply a func- generally expects to find decreased viru- tion of which mode of transmission that pre- lence in a horizontally-transmitted disease dominates. Under some circumstances, (Lipsitch et al., 1995a). Temporal or spatial increases in horizontal transmission may structure of the host population that results even decrease virulence (Lipsitch et al., in repeated or long-term contact with other 1995a). It is also clear that if parasites lower members of the population increases oppor- the fitness of the host, they will not survive tunities for parasite transmission without unless there is also some degree of hori- increased virulence. zontal transmission (Lipsitch et al., 1995b). (5) Novel hosts. Pathogens introduced Clearly, the of virulence in a into a novel host population may be partic- pathogen depends on complex interactions ularly virulent. This virulence may decrease between the host and the parasite; the trade- over time as in the classic example of myx- off between virulence and mode of trans- oma introduced to control rabbit pop- mission is but one important aspect. ulations in Australia. High virulence of the Other factors have also been suggested pathogen has been selected against because to modulate virulence: of the high host mortality (Fenner and (1) Vectored vs. directly-transmitted Ratcliff, 1965). pathogens. Vector-born pathogens are gen- (6) Pathogen replication rate. Increased erally more virulent than directly-transmit- replication rate increases host parasite load ted pathogens (Ewald, 1994). In general we and is associated with high virulence. would expect the adverse effects of high Within-host competition with other parasite virulence on the host to diminish opportu- strains may select for high replication rates, nities for parasite transmission to new hosts. and high virulence even if this diminishes However, this effect can be offset by the host survival and limits opportunities for vector, if it promotes parasite transmission. pathogen transmission (May and Anderson, (An example is . Though this dis- 1990). ease may incapacitate its host, the mobility (7) Life span of pathogen propagules. of the mosquito vector provides ample Virulence is favored in pathogens with long- opportunities for the malaria parasite to lived propagules provided the host-parasite reach new hosts). system is not in an equilibrium (Ewald, (2) Host density. Low host density favors 1987). In particular, in parasites that have low virulence whereas high host density recently invaded a susceptible host popula- favors high virulence (Bull, 1994). (High tion, greater propagule longevity may ini- host density increases transmission oppor- tially favour higher virulence; but once the tunities and thus offsets the tendency for equilibrium is reached the high virulence to lower transmission oppor- is independent of propagule longevity tunities). (Bonhoeffer et al., 1996). 202 I. Fries, S. Camazine

The field of evolutionary epidemiology is worker. However, from the point of view rapidly expanding. A substantial number of of parasite fitness, this form of vertical trans- theoretical papers have appeared in recent mission is of no greater consequence than years attempting to provide a coherent the horizontal worker to worker framework for understanding the evolution that occur with much greater frequency (due of disease virulence and its relationship to to the greater numbers of workers which factors such as the mode of parasite trans- interact with other workers in the colony, mission, host density and host behavior. compared to the interactions between the However, the perspectives provided by evo- solitary queen and her workers). (Note that lutionary epidemiology theories have only we do not consider the other potential form recently been applied to social insects. The of intracolony vertical transmission, transo- analysis by Schmid-Hempel (1998) covers varial infection, since this is not known to parasites in social insects in general, and occur in honey bees.) has some discussion of honey bees. How- In the following section we discuss hor- ever, only a few experimental studies of izontal and vertical pathogen transmission in social insects have direct bearing on evolu- greater detail, and the relative importance tionary epidemiology and transmission of of each of these modes of transmission for parasites, among them the studies of bumble various honey bee diseases. bees (Schmid-Hempel and Loosli 1998, Schmid-Hempel and Schmid-Hempel 1993, see also Schmid-Hempel 1998). 3. HONEY BEE SOCIETIES Modes of pathogen transmission: hori- AND DISEASE TRANSMISSION zontal and vertical When considering colonies of social As a relatively homogeneous community insects, we can subdivide horizontal and of thousands of related individuals, one vertical transmission into both intracolony would expect a honey bee colony to be (between individuals within a colony) and highly susceptible to rapid epidemic trans- intercolony (that between individuals from mission of disease organisms. The high den- different colonies) components (Tab. I). For sity of individuals within the colony, their example, horizontal intracolony pathogen close physical contact with one another transmission most commonly occurs when (through casual contact, communication, an infected worker bee transmits a disease to and mutual grooming) and the trophallac- another worker or a larva. Vertical intra- tic interchange of food and glandular sub- colony transmission occurs when an infected stances all provide numerous and diverse queen transmits a pathogen to a daughter opportunities for pathogen transmission.

Table I. Modes of pathogen transmission within and between honey bee colonies.

Horizontal Vertical

Intracolony Worker to brood, worker, or drone Queen to daughter (worker) Drone to worker or drone Queen to daughter (queen) Queen to son (drone)

Intercolony Worker to worker or drone Swarming Drone to worker or drone (drifting, robbing) Honeybee epidemiology 203

Furthermore the nest cavity, with its rela- 1. The pathogen must infect an individual tively constant temperature and moderate (and usually must be able to multiply within humidity, provides an ideal environment for this new host). pathogen growth and survival. 2. Generally, the pathogen must be able As a defense against pathogen transmis- to infect additional individuals within the sion within honey bee colonies, we find that colony to assure a sufficient parasite load honey bee colonies possess a number of within the colony. mechanisms that may have evolved specif- ically to curtail rampant spread of disease. 3. The pathogen must successfully gain These include: access to new colonies. (1) Use of antibacterial substances such as propolis to line the colony walls and seal In terms of fitness, the successful transfer the nest cavity; of a pathogen’s offspring to a new colony is a critical step in its life history. If a parasite (2) Antibiotic systems in pollen and or pathogen fails to achieve a foothold in honey stores; another host colony, the parasite will not (3) Hygienic cell cleaning behavior of increase its reproductive fitness, regardless the workers (reviewed in Boecking and of how prolific it has been within the origi- Spivak, 1999); nal host colony. Thus, hurdles #1 and #2 (4) An undertaker caste of bees which (intra-individual and intra-colony transmis- removes dead, potentially diseased, car- sion) are important aspects of pathogen fit- casses (Visscher 1983); ness only to the extent that they contribute to more efficient inter-colony transmission. (5) Grooming and allogrooming; (6) Physiological mechanisms such as In contrast, a human parasite faces only proventriculus that filters AFB spores, gut the two first hurdles. Consider, for exam- pH and immune mechanisms directed at ple, two intestinal parasites, the protozoan preventing infection. Entamoeba histolytica, infecting humans, and Nosema apis, a protozoan infecting the Although there are many characteristics epithelial cells of the honey bee ventricu- that limit intracolony parasite transmission, lus. Both diseases are associated with sim- few, if any, mechanisms have been ilar symptoms of intestinal dysentery. Enta- described whereby honey bee colonies moeba assures its fitness by infecting and restrict transmission of parasites between multiplying within its human host, and then colonies. The fact that foreign bees are using a fecal-oral route of transmission to rejected at the hive entrance during periods infect new hosts. Nosema does the same, of dearth could be associated with defense but is faced with the additional hurdle of against parasites, although this remains to gaining access to new colonies to assure its be verified. Even so, foreign bees loaded continued survival. It could be argued that with nectar are readily accepted during peri- humans also live in “colonies” (e.g. towns), ods of food abundance (Jay, 1965; Pfeiffer and that the third hurdle discussed above and Crailsheim, 1998). This indicates that applies also to humans. There may be some guarding as a mean of reducing parasite validity to this argument under circum- transmission is only partial, if it exists at all stances where groups of humans are com- for this purpose. paratively isolated. However, under normal From the disease organism’s viewpoint, circumstances, this is of minor importance, a honey bee pathogen must overcome three unlike the situation for honey bees in which distinct fitness hurdles in order to reproduce the colony is the predominant unit of selec- and disperse to new hosts: tion. 204 I. Fries, S. Camazine

3.1. Horizontal intracolony disease (4) Contact with infectious material from transmission the environment. The drifting of bees into the wrong Once a pathogen enters a colony or colony occurs frequently (Jay, 1965; Pfeiffer infects a bee, it may spread to other colony and Crailsheim, 1998) in apiaries where members. Such horizontal intracolony trans- colony densities are greater than under nat- mission is influenced by a variety of ural conditions. In contrast, there is little or pathogen life-history characteristics. Some no evidence for disease transmission by diseases affect only larval and pupal stages, drifting of individuals between feral colonies and do not produce disease in adult bees. In in the wild. In most cases, colonies are such cases, adult bees (often the nurse bees widely separated, precluding drifting of bees which feed the brood) may act as mechani- from one colony to another. cal vectors, carrying the disease to larvae. In other cases, the adult bees may become Robbing is a route of disease transmis- infected (but may show few or no symp- sion that probably occurs at significant lev- toms) and spread the disease to brood (e.g. els both under managed and natural condi- ) or other adults (e.g., nosema or tions. However, bees generally rob only amoeba disease where pathogens are trans- when there is little available foraging oppor- mitted through infected feces). Due to the tunities in the field, and they are only able to confinement of the bee larvae to their natal invade weak colonies. When outside food cells, disease transmission between the lar- sources become scarce, guard bees in strong val stages of honey bees is restricted to colonies usually detect and repel intruding infections where pathogens produce resting bees from other colonies. On the other hand, stages. When such resting stages are not suf- when colonies become diseased and weak- ficiently cleaned out from infected cells, the ened, guarding becomes ineffective and rob- next larva to occupy the cell may become bing bees easily enter a sick colony where infected. Infection can also occur when bees they may encounter pathogens. A robber are cleaning out cells and transmit infec- bee brings pathogens back to its own nest tious resting stages of the pathogens from on the surface of its body, or in robbed one cell to another. honey stored in its crop. An infected rob- ber could also infect the visited colony with pathogens on its body, although this route of 3.2. Horizontal intercolony infection seems less likely. transmission Generally, bees are unlikely to become infected during contact with another bee Horizontal intercolony transmission during foraging. However, parasite trans- of pathogens can occur through a number of mission via flowers has been reported routes. These include: (Kevan et al., 1990) and may be of some (1) Introduction of pathogens into a importance for the transmission of spiro- colony when an infected bee drifts from its plasma (Clark, 1977). own colony to another. There is the possibility that diseases such (2) Contact between individuals (or as chalkbrood and nosema may be spread between individuals and infectious materi- indirectly from individual to individual als) during robbing. Robbing is an activity in through contaminated water or other which foraging bees from one colony invade inanimate materials. For example, the another colony to steal honey. microsporidium Nosema apis can be (3) Contact between infected and unin- acquired by individuals during collection fected individuals from different colonies of water contaminated by bee feces during foraging. (L’Arrivée, 1965). Honeybee epidemiology 205

3.3. Vertical intercolony transmission additional (afterswarms) are headed by a virgin, yet unmated, queen. Intercolony vertical transmission is widespread, occur- Vertical transmission of parasites occurs ing whenever an infected colony swarms with the transfer of parasites between parents and brings with it infected individuals. and their offspring, from one host genera- Although some pathogens only affect brood, tion to the next (Fine, 1975). In the case of they can still be transmitted with adult bees. social insect colonies such as honey bees, Thus, there is an element of vertical trans- the most important form of vertical trans- mission involved in all honey bee diseases mission of pathogens occurs during repro- (see Tab. II). ductive swarming, when the colony divides and propagules (swarms) bud off from the parent colony. These new swarms are typi- 3.4. Disease virulence and pathogen cally headed by the old queen who leaves transmission in honey bees the original nest cavity to a daughter queen. In some cases a colony may produce sev- Here, we focus on the issue of why cer- eral swarms at a time, and in this case, the tain honey bee diseases are more virulent

Table II. List of common honeybee pathogens, trivial names, mode of intercolony transmission and virulence. + or +++ under transmission indicates which mode of transmission that is estimated to be of most importance for molding the host-parasite relationship.

Transmission

Type of Name of Trivial name of Horizontal Vertical Virulence pathogen pathogen disease/pathogen

Protozoa Nosema Nosema disease + +++ Benign apis Malphigamoeba Amoeba disease + +++ Benign mellificae

Fungi Ascosphaera Chalkbrood + +++ Benign apis Aspergillus flavus Stonebrood + +++ Benign

Bacteria Paenibacillus American +++ + Lethal larvae larvae foulbrood Melissococcus European + +++ Benign pluton foulbrood

Virus APV Acute paralysis virus + +++ Benign* DWV Deformed wing virus + +++ Benign*

Mites Acarapis woodi Tracheal mite + +++ Benign to lethal** Varroa destructor**** Varroa mite + +++ Benign to lethal***

* Only severe effects when vectored by Varroa mites. ** Only severe where the mite has been recently introduced. *** Only severe where the mite has been recently introduced or where effective mite control is employed. **** New species described from V. jacobsoni (Anderson and Trueman, 2000). 206 I. Fries, S. Camazine than others, and how apicultural practices cases of mite infestations are discussed sep- may affect disease virulence. We propose arately below, since virulence in newly intro- that the distinction between within-colony duced parasites requires special considera- virulence and between-colony virulence will tions (Anderson and May, 1986). have an important effect on the evolution of pathogen virulence in honey bees. At the colony level, we find that most diseases have low virulence. Whether this Disease virulence and pathogen trans- represents an adaptation of the parasite mission in colonies of honey bees present whose fitness interests are aligned with the a special case because host fitness depends host or an adaptive host response (or both) upon the ability of the colony to produce is difficult to determine. At the colony level, swarms. In turn, parasite fitness depends the only disease that has remained highly not only on the ability to transmit propagules virulent for many years is American foul- within the colony from one individual bee to brood (AFB). This unusual exception another, but also on its ability to move from requires an explanation and is therefore dis- one colony to another, either vertically, cussed in detail below. The only other dis- through swarming, or horizontally, through eases that can be considered virulent at bees and pathogens moving between colony level are the newly introduced mite colonies. diseases (discussed in the section on case studies). To understand honey bee epidemiology and disease virulence, one must distinguish A pathogen may be virulent within the between differences in pathogen transmis- colony, rapidly killing individual larvae, but sion and virulence at both the individual and non-virulent at colony level. Chalkbrood, colony level. Among individual bees, certain for example, is horizontally transmitted bacterial, fungal and viral diseases have high within the colony as adult bees carry spores intracolony virulence for larvae. In such that infect developing larvae. Once infected, cases, pathogens are transmitted to larvae the larvae quickly die. However, at the (from nurse bees, or through brood cell con- colony level, this pathogen is not virulent; tamination) and are highly virulent, rapidly colonies rarely, if ever, succumb to chalk- killing the larva. In contrast, high intracolony brood and although quantitative data is lack- virulence is almost unknown for adults. ing, we believe infected colonies often N. apis and Malpighamoeba mellificae can swarm and produce new colonies that carry be picked up by adult bees from contami- infectious spores on the swarming bees. In nated feces in the colony, but these para- the competition between different strains of sites are not especially virulent, and only chalkbrood within colonies, the more viru- modestly shorten a bee’s lifespan. In con- lent pathogen produces more propagules, trast, deformed wing virus (DWV) is a vir- and thus has the advantage. At the colony ulent disease that severely compromises level, on the other hand, high within-colony adult survival since bees with deformed virulence will be selected against if it wings cannot fly, and are readily ejected impedes colony reproduction. The predicted from the colony by other workers. Other outcome in this situation is a pathogen that viral diseases also cause high adult mortal- is virulent within the colony but not viru- ity, but all these viral diseases are related lent enough to keep colonies from swarm- to infestations with the parasitic mite, Var- ing. When vertical transmission of roa (Varroa jacobsoni, sensu lato) (Allen pathogens can be accomplished effectively and Ball, 1996). Tracheal mite (Acarapis with low virulence, then strains using hori- woodi) infestations can also result in high zontal transmission and high virulence will adult mortality, but this occurs only in over- be outcompeted (Lipsitch et al., 1996). Obvi- wintering bees (Gary et al., 1989). These ously, vertical transmission of pathogens Honeybee epidemiology 207 with low virulence can easily be achieved in (Rennie et al., 1921). When first discovered honey bees as colonies divide during swarm- they apparently caused considerable colony ing. If vertical transmission requires high losses, these mites are now considered a virulence, on the other hand, then strains of benign disease, and do not require control by much higher virulence and horizontal trans- beekeepers (Ruttner, 1971). In Finland, mission will be selected for (Lipsitch et al., where the mite was recently introduced, it is 1996). This second scenario is, we believe, spreading and causing considerable colony applicable to honey bees for one disease losses (Korpela, 1998). In North America, only – American foulbrood – with reasons the tracheal mite was first found in 1984 for this developed later. (Delfinado-Baker, 1984) and caused mas- An element of vertical parasite trans- sive colony losses (Scott-Dupree and Otis, mission is always present in honey bee 1990). Following the initial devastating colonies, since bees increase in numbers by effects from the tracheal mite infestations, dividing the mother colony to produce the dramatic impact of the parasite seems daughter colonies. Since vertical transmis- to have diminished and we expect this devel- sion of parasites favors development of a opment to continue. The hypothesis that the benevolent host-parasite relationship, highly mode of transmission molds pathogen vir- virulent parasites are unlikely to evolve in ulence predicts that a parasite such as the honey bees. Actually, swarming is likely to tracheal mite will be transmitted horizon- be advantageous for both the colony (since tally between colonies during the initial it reduces the parasite load within the phase before the infestation is well estab- colony) and for the parasite (since it pro- lished. Later, the vertical spread will be the vides a mechanism of disease transmission main source of transfer to new colonies as to new colonies). This regulatory effect of the parasite has become widespread. Over swarming has been proposed for honey bee time, the most susceptible colonies die and tracheal mites (Royce et al., 1991). Fur- the selective pressure on the parasite will thermore, evolution will likely favor vertical select against effects on colonies that inhibit transmission since it provides a safe route of reproductive swarming. Thus, a benign host transmission for pathogens. The parasites parasite relationship should evolve. can transfer from mother colonies to daugh- In the case of Varroa, which is a world- ter colonies, without ever leaving their hosts wide menace to beekeeping, we believe api- and facing exposure to the hazards of a hos- cultural practices are responsible for main- tile environment. taining virulent forms of the pathogen. In Table II summarizes the mode of trans- areas where the parasite has been established mission and pathogenicity at the colony level for several decades in honey bee popula- for a number of important honey bee tions, without being controlled by beekeep- pathogens. In Table II we find only three ers, the parasite no longer is lethal to infested pathogens that commonly result in colony colonies. This is the case in South America mortality: American foulbrood, tracheal both for Africanized bees and bees of Euro- mites, and Varroa mites. We discuss these pean origin (Rosenkranz, 1999) as well as in diseases in detail in the next section. North Africa (Ritter, 1990). The mecha- nisms for Varroa mite tolerance seem to be different among different sub-populations 3.5. Case studies of honey bees and probably developed inde- pendently (Rosenkranz, 1999). Note, how- 3.5.1. Mite diseases ever, that the precise nature of the honey bee- Varroa host parasite relationship is Tracheal mites were reported in central complicated by the fact that this parasite is Europe in the early part of this century not one species, but a species complex 208 I. Fries, S. Camazine

(Anderson and Fuchs, 1998), with the mite (Shimanuki and Knox, 1994). Spore viabil- infesting European bees recently redescribed ity allows the pathogen to “wait” long peri- as Varroa destructor (Anderson and True- ods and still be assured of eventual trans- man, 2000). Under the influence of apicul- mission to new colonies. Even when a tural management practices that promote colony dies, there is ample opportunity for opportunities for horizontal transmission, a transmission of the bacteria to new colonies. more virulent host-parasite relationship This would not be the case for other bacte- should be retained. With a long history of ria, viruses or protozoal pathogens that do co-adaptation on its natural host (the Asian not form such durable spores. honey bee, Apis cerana), the Varroa mite is in fact a benign parasite, as expected for a Spore forming pathogens like P. larvae pathogen that is primarily vertically trans- larvae, are effectively spread by robbing mitted. The European and Asian honey bees bees (Ratnieks, 1992). Robbing occurs com- have very similar life histories and it seems monly in apiaries. It also occurs among likely that Varroa should develop a benign widely separated feral colonies since bees host parasite relation in European honey can readily use olfactory cues to locate dis- bees, if given the opportunity. tant hives. There is no evidence that colonies avoid robbing disease-infected hives, and 3.5.2. American foulbrood infected nest sites are not avoided when new nests are established (Ratnieks and American foulbrood (AFB) (caused by Nowakowski, 1989). the spore-forming bacterium Paenibacillus In addition to pathogen spread through larvae larvae) regularly kills honey bee robbing or the occupation by swarms of colonies under natural conditions. The dis- infected sites, there is the potential for hor- ease is often insidious, with symptoms dis- izontal spread of AFB by drifting bees. covered by the beekeeper only late in the However, this route is probably of minor disease process (Hansen and Brødsgaard, importance even in commercial beekeeping 1999; Ratnieks, 1992). Although some where colony density may be extremely high colonies are resistant to the disease, sus- (Goodwin et al., 1994), and even less so ceptible colonies succumb to the infection. under natural conditions where low colony What is unusual about this parasite’s life density hinders intercolony drifting. Like- history that could promote a virulent, lethal wise, there is the possibility of spores being infection? Under natural conditions, the dis- carried by adult bees during swarming, ease readily spreads horizontally when although this is unlikely to result in clini- weakened, infected colonies are robbed by cally diseased colonies. From feeding exper- bees from other colonies, or when new iments, we know that very high spore levels swarms occupy nestsites where the former must be fed to colonies to produce clinical colony has succumbed to AFB. In general, symptoms of AFB (Hansen et al., 1988). horizontally transmitted parasites are Large numbers of spores need to be in the expected to modulate their virulence to strike colony to produce infection, although indi- a balance between their own reproduction vidual young larvae may succumb to dis- and host mortality (Anderson and May, ease from low spore doses (Hoage and 1982). If host mortality increases pathogen Rothenbuhler, 1966). Only strong colonies transmission between hosts, virulent forms produce daughter colonies and as colonies of the pathogen will be selected for. A major are weakened by AFB, they will be less difference between AFB and other honey likely to swarm, thus lowering the oppor- bee diseases is the presence of a long-lived tunity for vertical parasite transmission. spore that can retain its viability for decades Beekeeping practices to control AFB pro- under diverse environmental conditions vide further evidence that vertical pathogen Honeybee epidemiology 209 transmission during swarming is of minor have an impact on the development of importance for this disease. To cure foul- pathogen virulence since apicultural prac- brood-infected colonies, beekeepers shake tices alter the probability for horizontal ver- the adult bees off the old diseased combs sus vertical transmission of pathogens. and force them to build new combs (Hansen and Brødsgaard, 1999). This treatment is Apicultural practices increase the risk of similar to natural swarming since the swarm drifting of infected bees between colonies. constructs new comb as it establishes the Not only do drones frequently enter foreign colony. colonies (Free, 1958), but substantial drifting also occurs of workers (Jay, 1965; Pfeiffer The life-history strategy of AFB is prob- and Crailsheim, 1998). Management of ably the result of a selection pressure in colonies during periods of nectar shortage which the comparatively high rate of hori- and feeding of colonies often induces rob- zontal transmission compared to the low bing behavior, which again may transfer level of vertical transmission favors the infected bees or infected food stores between development of a relatively aggressive par- colonies. Besides these increased risks for asite. The opportunity for horizontal trans- horizontal pathogen transmission, high bee mission could be important in explaining densities also result in resource competition why AFB is the only lethal honey bee dis- between colonies for pollen and nectar, ease where the host parasite relationship has which may reduce overall colony fitness been molded over evolutionary times. The (Bailey and Ball, 1991). Furthermore, man- extremely stable spores produced by P. lar- aging colonies to increase production often vae larvae adds a further selective advan- involves shifting of comb (brood combs as tage for the pathogen since weakened or well as other comb) between colonies. This killed host colonies can still serve as sources greatly increases the probability of hori- of disease transmission between colonies. zontal transmission of all types of pathogen that may be transmitted by contaminated wax, such as nosema (Bailey, 1953), AFB 4. EPIDEMIOLOGICAL (Ratnieks, 1992) or chalkbrood (Koenig, IMPLICATIONS OF APICULTURAL 1987). Managing colonies also involves PRACTICES FOR THE moving bees between colonies (for exam- TRANSMISSION OF BEE DISEASES ple to strengthen weak colonies). This also facilitates horizontal transfer of disease 4.1. Effects of apicultural practices agents on or in adult bees. Beekeeping practices can have impor- As described earlier, honey bee colonies tant epidemiological consequences. Under possess a variety of mechanisms to limit natural conditions honey bees are restricted disease transfer within the colony, but these to natural cavities, such as hollow trees and systems may be overwhelmed by certain colony density is low. In most parts of the beekeeping practices that affect intracolony world, the vast majority of bee colonies are transmission. When bees are crushed dur- kept and managed by beekeepers. Housing ing hive manipulations, pathogens may for colonies is supplied by the beekeeper spread in the colony as bees clean out the and the density of colonies is much higher dead remains. Similarly, disease may be than in nature. Standard beekeeping prac- transmitted through feeders when bees defe- tices, which include swarm control and api- cate in the food. Hive manipulations that aries with large numbers of colonies, will induce defecation of bees inside the hive inevitably increase the horizontal transmis- may also promote disease transmission. Fur- sion and decrease the vertical transmission thermore, colony splits or other hive manip- of pathogens. We predict that this is likely to ulations may alter the age structure of the 210 I. Fries, S. Camazine bees and disturb the natural task distribu- perspective forming of new colonies under tion among workers. This can affect managed conditions will promote horizon- hygienic behavior performance, important tal transmission of pathogens to new for controlling AFB and other brood dis- colonies rather than the vertical transmis- eases (Spivak and Gilliam, 1999a, b). sion that occurs under natural swarming conditions. Among horizontally transmitted diseases, the evolution of pathogen virulence is also affected by host density: high host densi- 4.2. Changes in management practices ties favor higher virulence (Bull, 1994). The that would be expected to diminish rationale is that with high host densities, the disease virulence and transmission pathogen can more easily reach new hosts, and offset the fitness disadvantages to the Modern evolutionary epidemiology sug- host that occur with highly virulent gests that behavioral practices can greatly pathogens. In contrast to apicultural condi- affect the future course and virulence of dis- tions, the normal situation is for bee colonies ease (Ewald, 1994). Undoubtedly, apicul- to be widely scattered at low densities. ture creates numerous conditions where hor- Honey bees are long-lived under favorable izontal pathogen transmission is favored conditions. Provided the pathogen is not over vertical transmission. Thus, theory sug- highly virulent, this affords continued oppor- gests that apiculture per se will select for tunities for horizontal pathogen transmis- more virulent honey bee pathogens. As a sion between colonies even when they are consequence, beekeepers could benefit if widely scattered. Thus, even if the trade-off they instituted simple practices that reduce between vertical and horizontal pathogen horizontal transmission. For example, the transmission is not considered, lower viru- size of apiaries can be limited, and colonies lence is expected for horizontally transmit- can be placed to minimize drifting. Fur- ted pathogens in honey bees under natural thermore, the transfer of bees and brood conditions compared to conditions created between colonies should be limited as by apiculture. should all practices that increase the risks Honey bees swarm to reproduce. Swarm- for within hive defecation of bees or crush- ing behavior results in vertical pathogen ing of bees. Considering vertical pathogen transmission. Since bees that swarm pro- transmission, it is impractical to suggest that duce less honey, beekeepers generally colonies should be allowed to swarm to restrict swarming through a variety of man- favor this mode of transmission. However, agement practices. To increase the number a similar effect may be accomplished by of colonies or to replace losses, beekeepers paying attention to pathogen impact when make nuclei from old colonies. Normally breeding stock is selected. Beekeepers these nuclei receive a developing queen cell should continually select strains of bees that or a mated queen from pre-selected stock show disease resistance. and, thus, prevent the formation of daughter What we describe are good beekeeping colonies genetically related to the colonies practices. The evolutionary perspective on from which nuclei were formed. In this sys- transmission of pathogens simply adds tem, pathogens may be transferred to daugh- another dimension to old arguments about ter colonies on adult bees and with brood the management of bees, and suggests that combs. However, this is not equivalent to more severe disease problems can be vertical pathogen transmission since daugh- expected to develop unless good beekeeping ter colonies will, for the most part, be unre- practices are maintained. The predictions lated to the colonies from which they were presented in this paper align modern evo- formed. On the contrary, from a selective lutionary epidemiology with honey bee Honeybee epidemiology 211 pathology. The practical implications of this l’épidémiologie de l’évolution, cela devrait perspective still remain to be verified in field généralement sélectionner des relations hôte- experiments. parasite bénignes. Pourtant, la transmission inter-colonies des agents pathogènes de l’abeille se fait principalement verticale- ACKNOWLEDGEMENTS ment par l’essaimage et ces maladies pré- sentent une virulence comparativement Detailed comments by Dr. Stefan Fuchs on faible. La seule grande exception est la loque earlier manuscript versions are highly appreci- américaine (AFB), causée par la bactérie ated. Funding for this work was provided, in part, Paenibacillus larvae larvae. Les spores de by Pennsylvania Department of Agriculture. l’AFB restent viables durant des décennies et, une fois établie, l’infection est très viru- Résumé – Implications, pour l’épidémio- lente et conduit facilement à la mort des logie de l’abeille domestique, de la trans- colonies infectées. Elle se transmet facile- mission horizontale et verticale des agents ment horizontalement aux autres colonies pathogènes. Des discussions récentes sur et P. larvae larvae est probablement le seul l’épidémiologie des maladies suggèrent que agent pathogène de l’abeille pour lequel la la distinction entre transmission horizontale transmission horizontale de la maladie et transmission verticale d’un agent patho- domine. gène joue un rôle crucial dans la compré- Sans aucun doute, l’apiculture crée des hension de la formation de la virulence conditions dans lesquelles la transmission d’une maladie au cours de l’évolution. Com- horizontale est favorisée aux dépens de la parée à la transmission horizontale, la trans- transmission verticale. Ainsi la théorie sug- mission verticale est plus susceptible de gère que l’apiculture en soi sélectionne les sélectionner une relation hôte-parasite agents pathogènes les plus virulents. En bénigne qu’une virulente. conséquence les apiculteurs pourraient La transmission et la virulence des agents gagner à réduire la transmission horizon- pathogènes chez les colonies eusociales tale en limitant la taille des ruchers et en d’insectes, telles que les abeilles, présente un disposant les colonies de façon à restreindre cas particulier car la vitalité d’un agent au minimum la dérive. En outre, le trans- pathogène dépend non seulement de sa capa- fert des abeilles et du couvain d’une colonie cité à infecter et à se répandre parmi les indi- à l’autre devrait également être réduit au vidus de la colonie, mais aussi de sa capacité minimum de même que toutes les pratiques à se répandre d’une colonie à l’autre. Chez qui augmentent les risques d’écrasement les abeilles, une telle transmission inter- d’abeilles ou de défécation des abeilles à colonies peut se faire soit horizontalement l’intérieur de la colonie. (par exemple par la dérive d’abeilles d’une En ce qui concerne la transmission verti- colonie à une autre ou par des abeilles péné- cale, il est peu réaliste de suggérer de laisser trant temporairement dans d’autres colonies essaimer les colonies pour favoriser ce mode pour dérober du miel), soit verticalement de transmission. Pourtant on pourrait obte- (lorsque la colonie se reproduit par essai- nir un effet analogue en prêtant attention mage). Contrairement aux conditions aux maladies lors de la sélection des colonies d’exploitation dans les ruchers, où la dérive éleveuses. Les agents pathogènes qui se et le pillage ont lieu fréquemment en raison manifestent par des symptômes seront défa- des fortes densités de population, en condi- vorisés si les apiculteurs ne produisent de tions naturelles les colonies sont beaucoup nouvelles colonies qu’à partir de colonies plus espacées et la transmission verticale indemnes des symptômes cliniques. est la voie principale pour l’infection de Nous décrivons ici de bonnes pratiques api- nouvelles colonies. Selon la théorie de coles. Mais l’introduction en apiculture 212 I. Fries, S. Camazine d’une perspective évolutionniste sur la trans- Verhältnissen die vertikale Verbreitung mission des agents pathogènes ajoute une durch den hohen Abstand zwischen den Völ- autre dimension aux vieux arguments et sug- kern der hauptsächliche Weg, auf dem gère que des problèmes pathologiques plus pathogene neue Völker befallen. Die evo- sérieux vont progressivement voir le jour, lutionäre epidemiologische Theorie führt à moins que de bonnes pratiques apicoles zu der Voraussage, dass dies generell eine ne soient maintenues. Les prévisions pré- Selektion in Richtung eher gutartiger Wirt- sentées dans cet article s’alignent sur l’épi- Parasitbeziehungen führen sollte. démiologie de l’évolution et la pathologie In der Tat werden Krankheitserreger über- de l’abeille. Il reste à vérifier dans des expé- wiegend vertikal über die Schwarmbildung riences de terrain les implications pratiques weitergegeben, und diese Bienenkrankhei- de cette perspective. ten sind von vergleichsweise geringer Viru- lenz. Die einzige gröβere Ausnahme bildet Apis mellifera/ épidémiologie / transmis- die durch das Bakterium Paenibacillus lar- sion agent pathogène / transmission vae larvae hervorgerufene Amerikanische horizontale / transmission verticale Faulbrut (AFB). AFB-Sporen können über Jahrzehnte überleben, einmal erfolgte Infek- tionen sind hochvirulent und können befal- Zusammenfassung – Bedeutung hori- lene Völker leicht zugrunde richten. Die zontaler und vertikaler Verbreitung von Krankheit wird leicht von Volk zu Volk Pathogenen für die Entwicklung von übertragen und ist wahrscheinlich die ein- Krankheiten der Honigbienen. Die neue- zige Bienenkrankheit, bei der der horizontale ren Erörterungen der Entwicklung von Übertragungsweg dominiert. Krankheiten legen nahe, dass die Unter- Ohne Zweifel schafft die Bienenhaltung scheidung zwischen horizontaler und ver- Bedingungen, durch die die horizontale tikaler Ausbreitung eines Pathogens einen Übertragung gegenüber der vertikalen Über- wesentlichen Schlüssel zum Verständnis der tragung begünstigt wird. Die Theorie weist Ausformung der Virulenz einer Krankheit in darauf hin, dass Bienenhaltung an sich evolutiven Zeiträumen bietet. Vertikale Ver- geeignet ist, Bienenkrankheiten mit höhe- breitung führt wesentlich wahrscheinlicher rer Virulenz herauszuselektieren. Als zur Entwicklung gutartiger Wirt-Parasitbe- Schlussfolgerung hieraus könnte die Bie- ziehungen als horizontale Verbreitung. Die nenhaltung daraus Nutzen ziehen, die hori- Übertragung und Virulenz von Krankheit- zontale Weitergabe von Krankheiten durch serregern bei sozialen Insekten wie Honig- geringere Gröβe der Bienenstände und ver- bienen stellt einen besonderen Fall dar, da flugsmindernde Aufstellung der Völker zu die Lebensfähigkeit eines Parasiten nicht verringern. Weiterhin sollte der Austausch nur von der Verbreitung zwischen den Indi- von Bienen und Brut zwischen Völkern viduen innerhalb der Völker abhängt, son- sowie alle Vorgehensweisen vermieden wer- dern ebenso von der Fähigkeit der Verbrei- den, die das Abkoten in den Völkern oder tung zwischen Völkern. Bei Honigbienen das Zerdrücken von Bienen hervorrufen. kann eine solche Verbreitung ebenfalls hori- In Betrachtung vertikaler Weitergabe ist es zontal geschehen (etwa durch Verflug von praxisfern, zu verlangen, dass den Völkern Arbeiterinnen zwischen Völkern oder durch das Schwärmen ermöglicht werden sollte kurzzeitiges Eindringen um Honig zu steh- um diesen Übertragungsweg zu begünsti- len) oder aber vertikal, wenn sich die Völker gen. Allerdings könnte ein ähnlicher Effekt durch Schwärme vermehren. Im Gegensatz erreicht werden, wenn bei der Selektion von zu bewirtschafteten Völkern (bei denen Ver- Zuchtvölkern auf Krankheiten geachtet wird. flug oder Räuberei wegen der hohen Völ- Krankheitserreger, die symptomatische kerdichte häufig sind) ist unter natürlichen Effekte hervorrufen werden benachteiligt, Honeybee epidemiology 213 wenn die Bienenhalter neue Völker nur aus Bonhoeffer S., Lenski R.E., Ebert D. (1996) The curse von klinischen Symptomen freien Völkern of the pharaoh: The evolution of virulence in pathogens with long living propagules, Proc. R. erzeugen. Soc. Lond. B 263, 715–721. Was wir beschreiben ist gute Bienenhal- Bull J.J. (1994) Perspective: Virulence, Evolution 48, tungspraxis. Allerdings wird durch die evo- 1423–1437. lutionäre Perspektive auf die Rolle der Über- Bull J.J., Molineux I.J., Rice W.R. (1991) Selection of benevolence in a host-parasite system, Evolu- tragungswege von Krankheiten eine neue tion 45, 875–882. Dimension zu den alten Argumenten hin- Canning E.U. (1982) An evaluation of protozoal char- zugefügt, insbesondere da diese die Her- acteristics in relation to biological control of pests, ausbildung von immer schlimmeren Krank- Parasitology 84, 119–149. heitsformen vorhersagt, falls nicht gute Clark T.B. (1977) Spiroplasma as a new pathogen in the Bienenhaltungstechniken eingehalten wer- honeybees, J. Invertebr. Pathol. 29, 112–113. Delfinado-Baker M. (1984) Acarapis woodi in the den. Die in diesem Beitrag dargelegten Vor- United states, Am. Bee J. 124, 805–806. hersagen bringen moderne evolutionäre Epi- Ewald P.W. (1987) Transmission modes and evolu- demiologie mit der Honigbienenpathologie tion of the -mutualism continuum, Ann. auf Verbindung, es steht aber noch an, die N.Y. Acad. Sci. 503, 295–306. praktischen Folgerungen dieser Perspektive Ewald P.W. (1994) Evolution of infectious disease, in Feldversuchen zu Überprüfen. Oxford University Press, Oxford. Fenner F., Ratcliff F.N. (1965) Myxomatosis, Cam- bridge University Press, Cambridge, UK. Honigbienen / Epidemiologie / Krank- Fine P.E.M. (1975) Vectors and vertical transmission: heitsübertragung / horizontale Übertra- An epidemiologic perspective, Ann. N.Y. Acad. gung / vertikale Übertragung Sci. 266, 173–194. Free J.B. (1958) The drifting of honey bees, J. Agric. Sci. 51, 294–306. Gary N.E., Page R.E. Jr., Lorenzen K. (1989) Effect of REFERENCES age of worker honey bee (Apis mellifera) on tra- cheal mite (Acarapis woodi) infestation, Exp. Appl. Allen M., Ball B. (1996) The incidence and world dis- Acarol. 7, 153–160. tribution of honey bee viruses, Bee World 77, Goodwin R.M., Perry J.H., Ten-Houten A. (1994) The 141–162. effect of drifting honey bees on the spread of Amer- Anderson D.L., Trueman, J.W.H. (2000) Varroa jacob- ican foulbrood infections, J. Apic. Res. 33, 209–212. soni (Acari: Varroidae) is more than one species, Hafner M.S., Sudman P.D., Villablanca F.X., Spradling Exp. Appl. Acarol. 24, 165–189. T.A., Demastes J.W., Nadler S.A. (1994) Disparate Anderson D.L., Fuchs S. (1998) Two genetically dis- rates of molecular evolution in cospeciating hosts tinct populations of Varroa jacobsoni with con- and parasites, Science 265, 1087–1090. trasting reproductive abilities on Apis mellifera, Hansen H., Brødsgaard C.J. (1999) American foul- J. Apic. Res. 37, 69–78. brood: a review of its biology, diagnosis and con- Anderson R.M., May R.M. (1982) Co-evolution of trol, Bee World 80, 5–23. hosts and parasites, Parasitology 85, 411–426. Hansen H., Rasmussen B., Christensen F. (1988) A Anderson R.M., May R.M. (1986) The invasion, per- preliminary experiment involving induced infec- sistence and spread of infectious diseases within tion from Bacillus larvae, Dan. J. Plant Soil Sci. animal and plant communities, Philos. Trans. R. 92, 11–15. Soc. Lond. B 314, 533–570. Herre E.A. (1993) Population structure and the evo- lution of virulence in nematode parasites of fig Antia R., Levin B.R., May R.M. (1994) Within-host wasps, Science 259, 1442–1445. population dynamics and the evolution and main- tenance of microparasite virulence, Am. Nat. 144, Hoage T.R., Rothenbuhler W.C. (1966) Larval honey 457–472. bee response to various doses of Bacillus larvae spores, J. Econ. Entomol. 59, 42–45. Bailey L. (1953) The transmission of nosema disease, Hurd H., Lane R. (1998) Parasite interactions: recip- Bee World 34, 171–172. rocal manipulation, Symposia of the British Soci- Bailey L., Ball B.V. (1991) Honey bee pathology, Aca- ety for Parasitology, Vol. 35, Parasitology 115 sup- demic Press, London. plement 1998, pp. 1–115. Boecking O., Spivak M. (1999) Behavioral defenses of Jay S.C. (1965) Drifting of honeybees in commercial honey bees against Varroa jacobsoni Oud., Api- apiaries. 1. Effect of various environmental fac- dologie 30, 141–158. tors, J. Apic. Res. 4, 167–175. 214 I. Fries, S. Camazine

Kevan P.G., Laverty T.M., Denmark H.A. (1990) Asso- Royce L.A., Rossignol P.A., Burgett D.M., Stringer ciation of Varroa jacobsoni with organisms other B.A. (1991) Reduction of tracheal mite parasitism than honeybees and implications for its dispersal, of honey bees by swarming, Philos. Trans. R. Soc. Bee World 71, 119–120. Lond. B 331, 123–129. Koenig J.P. (1987) Factors contributing to the patho- Ruttner F. (1971) Milben-Panik, Die Biene 107, genesis of chalk brood disease in honey bee 201–203. colonies, Ph.D. thesis, University of Wisconsin, USA. Schmid-Hempel P. (1995) Parasites and social insects, Apidologie 26, 255–271. Korpela S.L. (1998) Pest status and incidence of the honey bee tracheal mite in Finland, Agric. Food Schmid-Hempel P. (1998) Parasites in social insects, Sci. Finland 7, 469–476. Princeton University Press, Princeton, New Jersey. L’Arrivée J.C.M. (1965) Sources of nosema infection, Schmid-Hempel P., Loosli R. (1998) A contribution Am. Bee J. 105, 246–248. to the knowledge of Nosema infections in bumble Lipsitch M., Herre E.A., Nowak M.A. (1995a) Host bees Bombus spp., Apidologie 29, 525–535. population structure and the evolution of virulence: Schmid-Hempel P., Schmid-Hempel R. (1993) Trans- a “law of diminishing returns”, Evolution 49, mission of a pathogen in Bombus terrestris, with a 743–748. note on division of labour in social insects, Behav. Lipsitch M., Nowak M.A., Ebert D., May R.M. (1995b) Ecol. Sociobiol. 33, 319–327. The population dynamics of vertically and hori- zontally transmitted parasites, Proc. R. Soc. Lond. Scott-Dupree C.D., Otis G.W. (1990) The impact of B 260, 321–327. Acarapis woodi (Rennie), the honey bee tracheal mite, on honey bees (Apis mellifera L.) in New Lipsitch M., Siller S., Nowak M.A. (1996) The evolu- York State, in: Proc. Int. Symp. Recent Res. Bee tion of virulence in pathogens with vertical and Pathol., 1990, Janssen Pharmaceutica, Beerse, Bel- horizontal transmission, Evolution 50, 1729–1741. gium, 5–7 Sept. 1990, Gent, Belgium, pp. 108–115. May R.M., Anderson R.M. (1990) Parasite-host coevo- lution, Parasitology 100, 89–101. Shimanuki H., Knox D.A. (1994) Susceptibility of Bacil- lus larvae to Terramycin®, Am. Bee J. 134, 125–126. Moritz R.F.A., Southwick E.E. (1992) Bees as super- organisms, Springer Verlag, Berlin. Shykoff J.A., Schmid-Hempel P. (1991) Parasites and Pfeiffer K.J., Crailsheim K. (1998) Drifting of honey- the advantage of genetic variability within social bees, Insectes Soc. 45, 151–167. insect colonies, Proc. R. Soc. Lond. B 243, 55–58. Ratnieks F.L.W. (1992) American foulbrood: the spread Spivak M., Gilliam M. (1998) Hygienic behaviour of and control of an important disease of the honey honey bees and its application for control of brood bee, Bee World 73, 177–1914. diseases and Varroa. Part I. Hygienic behaviour Ratnieks F.L.W., Nowakowski J. (1989) Honeybee and resistance to American foulbrood, Bee World swarms accept bait hives contaminated with Amer- 79, 124–134. ican foulbrood disease, Ecol. Entomol. 14, 475–478. Spivak M., Gilliam M. (1998) Hygienic behaviour of Rennie J., White P.B., Harvey E.J. (1921) Isle of Wight honey bees and its application for control of brood disease in hive bees, Trans. R. Soc. Edinb. 52, diseases and Varroa. Part II. Studies on hygienic 737–779. behaviour since the Rothenbuhler era, Bee World Ritter W. (1990) Entwicklung einer Varroatoleranz in 79, 169–186. Bienenvölkern in Tunesien, Allg. Dtsch. Imkerztg. Turner P.E., Cooper V.S., Lenski R.E. (1998) Tradeoff 24, 7–10. between horizontal and vertical modes of trans- Rosenkranz P. (1999) Honey bee (Apis mellifera L.) mission in bacterial plasmids, Evolution 52, tolerance to Varroa jacobsoni Oud. in South Amer- 315–329. ica, Apidologie 30, 159–172. Visscher P.K. (1983) The honey bee way of death: Royce L.A., Rossignol P.A. (1990) Epidemiology of necrophoric behaviour in Apis mellifera colonies, honey bee parasites, Parasitol. Today 6, 348–353. Anim. Behav. 31, 1070–1076.

To access this journal online: www.edpsciences.org