How Does the Mite Varroa Destructor Kill the Honeybee Apis Mellifera?

Journal of Insect Physiology 58 (2012) 1548–1555

Contents lists available at SciVerse ScienceDirect

Journal of Insect Physiology

journal homepage: www.elsevier.com/locate/jinsphys

How does the mite Varroa destructor kill the honeybee Apis mellifera? Alteration of cuticular hydrcarbons and water loss in infested honeybees ⇑ Desiderato Annoscia, Fabio Del Piccolo, Francesco Nazzi

Dipartimento di Scienze Agrarie e Ambientali, Università degli Studi di Udine, Via delle Scienze 208, 33100 Udine, Italy article info abstract

Article history: Several factors threaten the health of honeybees; among them the parasitic mite Varroa destructor and Received 24 May 2012 the Deformed Wing Virus play a major role. Recently, the dangerous interplay between the mite and Received in revised form 20 September the virus was studied in detail and the transition, triggered by mite feeding, from a benign covert infec- 2012 tion to a devastating viral outbreak, characterized by an intense viral replication, associated with some Accepted 22 September 2012 characteristic symptoms, was described. In order to gain insight into the events preceding that crucial Available online 3 October 2012 transition we carried out standardized lab experiments aiming at studying the effects of parasitization in asymptomatic bees to establish a relationship between such effects and bee mortality. It appears that Keywords: parasitization alters the capacity of the honeybee to regulate water exchange; this, in turn, has severe Apis mellifera Cuticular hydrocarbons effects on bee survival. Varroa destructor These results are discussed in light of possible novel strategies aiming at mitigating the impact of the Water loss parasite on honeybee health. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction 2012), so that, during the Summer, all bees in a colony can be re- garded as DWV infected by different number of viral particles The mite Varroa destructor Anderson & Trueman is the most (Nazzi et al., 2012). important parasite of the western honeybee Apis mellifera L. In a previous study we showed that collapse in mite infested (Sammataro et al., 2000). Since it appeared in the Western World colonies occurs, at the end of the active season, because of a sud- it has caused huge damage to the beekeeping industry and has been den increase of bee mortality caused by lethal numbers of DWV linked to widespread losses of honeybee colonies recorded in the US genome copies resulting from the viral replication triggered by and Europe since 2006 (Le Conte et al., 2010). Recent studies mite infestation (Nazzi et al., 2012). However, the effects of mite showed that the Varroa mite plays a critical role in the collapse of infestation on individual honeybees, before the virus reaches lethal honeybee colonies. In particular, it has been shown that the mite levels and the symptoms of active viral replication (i.e. deformed is related to dramatic changes in strain diversity within Deformed wings) are noted, have not been investigated in detail yet. Wing Virus (DWV) that are likely related to increased prevalence Several studies have been carried out so far dealing with the and viral load (Martin et al., 2012). Other authors, investigating consequences of mite infestation on honeybee physiology and the dynamics of collapse leading interactions, showed that the mite health (Yang and Cox-Foster, 2007 and citations therein) but, due can destabilize the within-host dynamics of DWV with dramatic to the close relationship between the Varroa mite and DWV, the consequences on honeybee mortality (Nazzi et al., 2012). distinction of the causal agent of the observed effects is very difﬁ- DWV is a widespread honeybee virus normally causing benign cult. However, apart from its biological interest, the study of the ef- covert infections that, in presence of the mite, can turn into devas- fects of mite infestation in the absence of viral infection and tating viral outbreaks (de Miranda and Genersch, 2010). Overt viceversa is probably of limited value under a practical point of infections are characterized by high loads of viral genome copies view, given the ubiquitous distribution of both V. destructor and and some classic symptoms including crippled wings and short- DWV. Instead, the study of the effects of mite infestation on appar- ened abdomen (Möckel et al., 2011). Several studies demonstrated ently healthy, albeit infected bees and, moreover, the elucidation of that, in Europe, most honeybee colonies are infected (see Ribière the biological phenomena underlying the observed effects would et al., 2008 for a review) and prevalence increases throughout be vital in view of developing intervention strategies aiming at the year (Tentcheva et al., 2004; Martin et al., 2010; Dainat et al., mitigating the effects of mite infestation. The most apparent effect of mite infestation at the individual level appears to be weight loss (Yang and Cox-Foster, 2007 and cita- ⇑ Corresponding author. Tel.: +390432558513; fax: +390432558501. E-mail address: [email protected] (F. Nazzi). tions therein). Changes in the content of sugars and proteins

(Bowen-Walker and Gunn, 2001) have been reported in infested The experiment was repeated 6 times using in total 77, 68 and bees as well as alterations of the cuticular hydrocarbons (Salvy 52 bees that had been infested during the pupal stage with none, et al., 2001) but the very cause of the reduced weight still remains one or three mites respectively. For each replicate, new sets of cells unknown. infested by none, one or three mites were prepared; the figures re- In order to fill this gap, we carried out a study aiming at inves- ported above refer to the sum of cells of each type that were used. tigating in detail both the effects of Varroa parasitization on individual asymptomatic honeybees and the underlying causes. To 2.4. Influence of weight at the eclosion on the longevity of adult bees keep under control possible confounding factors that can affect field data, we opted for a lab study in which infested bee larvae In order to verify the potential influence of the weight on the were maintained under controlled conditions in artificial rearing longevity of adult honeybees, we studied the survival of individual cells. This method was used throughout the present study both bees that had been weighted at the eclosion. To do so, bee larvae to confirm and integrate available data on the outcome of mite were infested with none, one or three Varroa mites as described infestation and for a thorough investigation on the possible causes above, then, after 12 days, cells were opened and the emerging of the observed effects. adult bees were weighted using an electronic precision balance We believe that, besides the obvious biological interest in (Sartorius CP2P, maximum loading capacity 2000 mg, readability unraveling this aspect of the host-parasite interaction, a precise 0.001/0.002/0.005 mg), marked with a queen marker kit, consisting description of such effects is vital in view of developing an inte- of colored-numbered tags (2 mm Ø) applied on the thorax with a grated strategy to sustain honeybee health. resin type glue, and transferred into plastic cages as already described. At the emergence, symptoms of DWV infection such as de- 2. Materials and methods formed wings were noted if present; during the experiments, daily controls were carried out to note and remove any dead bee. 2.1. Biological material The experiment was repeated three times using in total 35, 28 and 22 bees that had been infested during the pupal stage with The biological material (larvae and adult females of V. destruc- none, one or three mites respectively. tor) was obtained from an experimental apiary located in Udine (North-eastern Italy). Previous studies indicated that local colonies 2.5. Honeybee water content are hybrids between Apis mellifera ligustica Spinola and Apis mellifera carnica Pollman (Comparini and Biasiolo, 1991; Nazzi, 1992). To assess water content of eclosing bees infested or not by one The apiary was set up in March and was left untreated for the mite, capsules used for rearing were opened, after 12 days, and the duration of the experiments; the biological material was collected honeybee fresh weight recorded, using an electronic precision bal- from April to September. ance, after removing the infesting mites and their offspring. Mites and the bee larvae were collected from brood cells capped Then samples were maintained at 80 °C for 24 h on a Petri dish 0–15 h previously (Nazzi et al., 2012). (Sage, 1982) and the dry weight of each honeybee registered again Previous studies showed that, in the area, the prevalence of after 30 min at room temperature. DWV is about 15% at the beginning of the Spring but it raises stea- The experiment was replicated 4 times using 50 honeybees for dily to about 100% by the middle of the Summer (Nazzi et al., each experimental group. 2012). Thus, in all cases, the vast majority, if not the total, of bees used in this work were infected by the virus as confirmed by PCR 2.6. Dynamics of weight-loss during pupation tests conducted in the framework of a parallel study.

Larvae not infested or infested with one mite, obtained as de- 2.2. Artiﬁcial infestation of bee pupae scribed above, were transferred into gelatin capsules (Assing, 6.5 mm Ø), weighted and then maintained in a climatic chamber Larvae obtained as above were transferred into gelatin capsules (34 °C, 75% R.H., dark) for 12 days. The weight of each larva was (Assing, 6.5 mm Ø) with none, one or three mites and maintained noted before and after encapsulation, in order to obtain the weight in a climatic chamber (34 °C, 75% R.H., dark) until the day of the of the empty capsule and the Varroa mite; then the weight of the eclosion, 12 days later (Nazzi and Milani, 1994). The cap of the cells capsules was noted twice a day (midmorning and midafternoon) was pierced 3 or 4 times using a No. 2 insect pin to make gas ex- during the entire pupation period. change with the exterior possible. Weights were registered using an electronic precision balance placed inside the climatic chamber. During the experiment, dead 2.3. Inﬂuence of mite infestation and viral infection on the survival of or abnormal pupae were discarded; after 12 days, cells were honeybee opened and the possible presence of deformed wing adult bees was noted. In total, 42 larvae (21 uninfested and 21 infested with After 12 days, capsules containing newly emerged adult bees one mite) were considered; out of these, 14 uninfested and 6 in- (either uninfested or infested with one or three mites) were fested bees were asymptomatic and were used for data analysis. opened and inspected under the microscope; the infesting mites and their offspring were separated from their host. Honeybees 2.7. Extraction of honeybee cuticular hydrocarbons belonging to the same experimental group were then transferred into a plastic cage (185 Â 105 Â 85 mm) in a climatic chamber Cuticular extracts of adult honeybees (either uninfested or in- (34 °C, 75% R.H., dark) and fed with sugar candy (Apifonda) and fested with one or three mites), obtained from larvae maintained water ad libitum; cages hosted ten to twenty bees at a time. in gelatin capsules for 12 days, were analyzed to evaluate the ef- During the experiments, daily controls were carried out to note fects of mite infestation on the composition of cuticular lipids. and remove any dead bee; possible symptoms of DWV infection Extracts for gas chromatography-mass spectrometry (GC–MS) such as deformed wings were noted if present; this allowed to cal- analysis were obtained by introducing a single bee into a conical culate the proportion of deformed wing bees in each experimental glass vial with 1 ml of hexane (Fluka Analytical) and 100 ng of 1- group for each replicate. octadecene (Aldrich Chemistry) as the internal standard (IS). After 1550 D. Annoscia et al. / Journal of Insect Physiology 58 (2012) 1548–1555

15 min, the extract was transferred into a smaller vial and reduced ad libitum. During the experiment, daily controls were carried out under nitrogen to 10 ll. One ll of the extract, corresponding to 0.1 to note and remove dead bees. bee equivalent, was injected in the GC–MS for the analysis. The experiment was replicated three times; in total, 40 uninfested and 37 infested bees, maintained at 60% R.H., and 42 uninfested and 33 infested bees, maintained at 75% R.H., were 2.8. GC–MS analysis considered.

The compounds present in the cuticular extracts of the honey- 2.10. Statistical analysis bees were identified with a Varian 3400 gas chromatograph cou- pled to a Varian Saturn 2000 mass spectrometer. The column The statistical tests listed below were used to analyze the re- (CP-SIL 8, 30 m Â 0.25 mm ID, film thickness: 0.25 lm) was main- sults of the experiments. tained at 50 °C for 1 min then programmed to 320 °Cat10°C/min. The percentages of bees with deformed wings among those in- The carrier gas was helium (flow: 1 ml/min). Injection volume was fested by none, one or three mites were compared using the Mar- 1 ll in splitless mode. MS ionization energy was set to 70 eV. Iden- ascuilo procedure. tification was based on the comparison of the mass spectra and the Average life span, weight at the eclosion and water content of retention time with those of authentic standards. Quantitative bees from different experimental groups were compared using analysis of the identified compounds was carried out with the General Linear Models; the Least Square Method was used for internal standard method, using 1-octadecene as the IS. In total, regression analyses. Comparison of mortality in bees exposed to the cuticular hydrocarbons of 5 bees per experimental group were different levels of infestation and different humidity regimes was extracted and analyzed; the same extract was analyzed 3–4 times carried out using the Chi-square test. and the average result used for subsequent analyses. For data anal- The Marascuilo procedure was applied using an on-line pro- ysis, only compounds with a relative peak area of more than 1% gram (http://www.stattools.net/index.php). The statistical soft- occurring in 60% of the members of at least one group were re- wares MiniTab (GLM), Statplus (regression analysis) and tained; this reduced the number of variables to 12. All peak areas Statistica (discriminant analysis) were used for all the other were standardized to 100% and transformed according to Reyment analyses. (1989). To reduce the number of variables and avoid multicolline- arity, we first performed a principal component analysis, then the extracted factors were used to separate the groups with a discrim- 3 Results and discussion inant analysis. 3.1. Mite infestation, wing deformity and survival

2.9. Effects of humidity on the survival of infested and uninfested The proportion of bees with the characteristic symptom of an honeybees overt virus infection (i.e. deformed wings) was higher in infested bees compared to uninfested ones and related to the infestation To corroborate the hypothesis that infestation affects the capac- rate (Overall Chi Square = 10.4242, df = 2: P = 0.0055; Fig. 1). ity of bees to reduce water loss through the cuticle, we carried out The small proportion of bees with deformed wings among the an experiment in which we exposed infested and uninfested bees uninfested ones indicates that mite infestation is a sufficient but to two different humidity regimes and compared mortality in the not necessary condition for the appearance of the symptoms of first week of life: the time interval when, according to data re- an overt virus infection and other causes may be involved. ported here, mite infestation appeared to cause the strongest In general, the results support previous findings about the effect effect. of mite infestation on viral replication (Yang and Cox-Foster, 2005, In this case, adult honeybees (either uninfested or infested with 2007; Möckel et al., 2011; Nazzi et al., 2012). Yang and Cox-Foster one mite) obtained from larvae maintained in gelatin capsules for (2005) first hypothesized that this may be the result of a direct 12 days were transferred into plastic cages, after the removal of the immunosuppressive action of the mite; a recent study suggests infesting mites, and maintained in the dark at 34 °C at 60% or 75% that the mite indirectly triggers viral replication by activating R.H.. Honeybees were fed with sugar candy (Apifonda) and water NF-jB dependent clotting and melanisation thus exploiting the

Fig. 2. Survival (days) of bees infested with none (L), one (L + 1 V) or three mites Fig. 1. Proportion of bees with deformed wings among those eclosed from pupae (L + 3 V) during the pupal stage. Data about normal and deformed wing bees are that were infested by none (L), one (L + 1 V) or three mites (L + 3 V) during the pupal reported separately. The error bars indicate the standard deviation. Deformed wing stage. The error bars indicate the standard deviation. The proportion of bees with bees emerging from larvae infested by none, one or three mites lived much shorter the characteristic symptom of an overt virus was higher in infested bees compared than corresponding normal wing bees. Mite infestation reduced the lifespan of to uninfested ones and related to the infestation rate. normal wing honeybees as well as that of deformed wing ones. D. Annoscia et al. / Journal of Insect Physiology 58 (2012) 1548–1555 1551

Fig. 3. Bee mortality of normal wing honeybees that were artiﬁcially infested with none (L), one (L + 1 V) or three mites (L + 3 V) during the pupal stage, in different time intervals. Infestation causes a dramatic change in the survival of honeybees; most infested bees died within two weeks of age.

wing bees could explain, at least in part, their reduced survival in that such bees can hardly feed due to the limited mobility; however further experiments, going beyond the aim of this study, would be needed to assess the very cause of the observed effect. Under field conditions, the aggressive behavior displayed by normal house bees towards deformed kins that was noted in an observation hive (unpublished observations) could significantly aggravate the impact of wing deformity on the survival of affected bees. Furthermore, mite infestation reduced the lifespan of normal wing honeybees as well as that of deformed wing ones (F = 18.22, df = 2: P < 0.001; Fig. 2), thus a significant effect of infestation on survival can also be observed in bees showing no symptoms of an overt virus infection. In this case, most infested adults Fig. 4. Average weight (mg) of honeybees that were infested with none (L), one died within two weeks of age, whereas uninfested ones showed lit- (L + 1 V) or three mites (L + 3 V) during the pupal stage. Data about normal and tle mortality in the same time interval (Fig. 3). In particular, a very deformed wing bees are reported separately. The error bars indicate the standard significant difference in mortality between infestation rates was deviation. The weight of infested honeybees, with both normal and deformed recorded in the first week of life (Overall Chi Square = 31.5819, wings, was significantly lower than that of uninfested ones; the weight reduction is related to the degree of infestation. df = 2: P < 0.001), suggesting that this is the period when infestation suffered during the pupal stage displays its negative effects on adult bee survival. limited cellular pool of a transcription factor that is involved in antiviral defense (Nazzi et al., 2012). 3.2. Weight and survival in infested honeybees Deformed wing bees emerging from larvae infested by none, one or three mites lived much shorter than corresponding normal The weight of infested honeybees, with both normal and de- wing bees (F = 44.49, df = 1: P < 0.001; Fig. 2). formed wings, was significantly lower than that of uninfested ones Thus an overt viral infection is strongly associated to a marked and the weight reduction seems to be related to the degree of reduction of honeybee lifespan. The physical deficit of deformed infestation (F = 36.6, df = 2: P < 0.001; Fig. 4).

Fig. 5. Correlation between weight (mg) and life span (days) in normal wing honeybees infested with one mite during the pupal stage. Bee weight at the eclosion and survival are correlated in normal wing single infested bees. 1552 D. Annoscia et al. / Journal of Insect Physiology 58 (2012) 1548–1555

Fig. 6. Dry weight (mg) and water content (%) in uninfested (L) and infested (L + 1 V) honeybees. The error bars indicate the standard deviation. The water content of infested bees was lower than that of uninfested bees, whereas the dry weight was not affected by the infestation.

On the other hand, deformed wing bees were not lighter than that, in turn, can affect survival. Therefore, we aimed at explaning normal wing ones with similar infestation rates (F = 1.68, df = 1: the causes of reduced weight of normal wing infested bees, the P = 0.199; Fig. 4), thus deformity itself does not seem to be related most abundant cohort of bees in an infested honeybee colony un- to weight reduction. der normal conditions. A significant correlation between bee weight at the eclosion and survival was observed in normal wing single infested bees 3.3. Cuticle and water balance in infested bees (y = 0.4325x À 32.026, R2 = 0.3722: P = 0.003; Fig. 5); a lower, not significant, correlation was observed in bees infested with three The water content of infested bees was lower than that of unin- mites (y = 0.1639x À 9.6155, R2 = 0.1278: P = 0.21), although, in fested bees (F = 25.57, df = 1: P < 0.001; Fig. 6), whereas the dry this case, the sample size was smaller. Instead, the relationship be- weight did not seem to be much affected by the infestation. Thus tween weight and lifespan was not noted in uninfested bees the reduced weight, observed in infested bees (Fig. 4), seems to (y = À0.0211x + 20.134, R2 = 0.0013: P = 0.853). be mostly related to a reduction in the water content of bees. This The reduced weight observed in infested bees confirm previous was already noted by Bowen-Walker and Gunn (2001), who sug- data showing that this is the most apparent effect of parasitization gested that the effect could be related to the subtraction of haemo- (Yang and Cox-Foster, 2007 and citations therein). The substantial lymph because of the feeding activity of the mite, as the weight similarity of weight in deformed and normal wing bees suggests reduction observed in infested bees is compatible with the esti- that weight loss is not strictly associated to an overt virus infection mated amount of haemolymph that a mite can subtract through but it is, instead, related to mite infestation. feeding during bee development. However, since under normal The significant relationship between weight and survival ob- conditions, transpiration through the cuticle is the main route of served in infested bees but not in the uninfested ones suggests that water loss in insects (Hadley, 1994), any possible effect of infesta- the weight per se is not related to survival but weight reduction tion on such process should be considered as well. This view is sup- represents a good proxy indicator of the intensity of mite damage ported by the results of an experiment in which the weight of

Fig. 7. Weight (mg) of a rearing capsule containing one uninfested (L) or one infested (L + 1 V) honeybee during pupation; the ratio between the weight of uninfested and infested capsules is reported as well as the equation of the regression line to highlight the different dynamics of weight change in the two groups. The weight of the rearing capsule containing an infested honeybee pupa together with the infesting mite changed more rapidly compared to that of the capsule containing an uninfested pupa. D. Annoscia et al. / Journal of Insect Physiology 58 (2012) 1548–1555 1553

factor involved in water balance in arthropods (Gibbs, 1998), this result points towards a possible mite induced alteration of the outer layer of the insect’s body surface, causing increased water loss and the related weight reduction. The amount of lipids that could be extracted from the cuticle of a newly emerged honeybee, either infested or not by the Varroa mite, was about 40 lg, confirming previous data (Francis et al., 1989). A GC–MS analysis of such extracts revealed the presence of several linear and branched saturated hydrocarbons with an odd number of carbon atoms from 19 to 33, as well as the corresponding monounsaturated hydrocarbons (Fig. 8) as already shown by several authors (Francis et al., 1989; Salvy et al., 2001; Del Piccolo et al., 2010). Overall the comparison between uninfested and infested bees revealed a substantial qualitative similarity (Fig. 8); however, a discriminant analysis of the cuticular hydrocarbons of bees infested by none, one or three mites revealed a clear difference in the percentage composition of cuticular extracts (the percentage of correct group classification being 100% for all groups). This result, obtained under artificial conditons, without any possible interference from the lipids coming from the nest, confirms the effect of infestation on the hydrocarburic layer covering the insect’s cuticle (Fig. 9) reported by Salvy et al. (2001). It appears that mite infestation during the pupal stage interferes with cuticle development, affecting the amount of some of the hydrocarbons that are deposited on the outer insect surface causing a shift in the relative proportions of such compounds. The observed effect may be related to the subtraction of haemolymph where lipophorin ligated hydrocarbons are released by oenocytes for being transported to the cuticle (Bagnères and Blomquist, 2010). On the other hand, in light of available data on mite-induced reduction of energetic compounds noted by other authors (Bo- wen-Walker and Gunn, 2001), we can speculate that, in case of shortage of carbohydrates, some compounds, that are used as pre- cursors for hydrocarbon biosynthesis (Blomquist, 2010), may be redirected to other pathways for being used in energetic metabo- lism. However, further studies (e.g. using radiolabeled hydrocarbons) are needed to support these arguments. Several studies have been carried out on the relationship between cuticular hydrocarbons composition and water balance in insects but a definitive conclusion has not yet been reached (Gibbs, 1998). In any case, it is likely that the alteration of the cuticular hydrocarbons observed in case of infestation may alter the permeability of the cuticle, affecting the capacity of the honeybee to prevent water loss from the body surface. Fig. 8. Cuticular profiles of honeybees infested by none (L), one (L + 1 V) or three mites (L + 3 V) during the pupal stage as from the GC–MS analysis of extracts. The profile of cuticular hydrocarbons of bees infested by none, one or three mites were 3.4. Effects of humidity on the survival of infested and uninfested qualitatively similar. honeybees

In the first week of observation, infested bees, exposed to lower infested bees was checked throughout development. In this case, relative humidity levels, showed a reduced survival compared to the weight of rearing capsules containing an infested honeybee those maintained at higher humidity (Chi square = 4.210, df = 1: pupa together with the infesting mite decreased more rapidly com- P = 0.04; Fig. 10) whereas uninfested bees did not suffer from the pared to that of the capsules containing an uninfested pupa (Fig. 7). lower humidity (Chi square = 0.2936, df = 1: P = 0.59; Fig. 10), con- This suggests the possibility that some factor other than the simple firming the relationship between infestation and water stress tol- haemolymph subtraction may account for the weight reduction erance. Thus, it appears likely that the observed alteration of observed in infested bees. In fact, although this may not be re- cuticular hydrocarbons that is induced by mite infestation, may af- garded as a conclusive evidence, if the weight reduction of the fect the capacity of the cuticle to prevent water loss and, in turn, bee pupae was solely due to haemolymph subtraction by the mite, the water stress tolerance of infested honeybees. Further than the weight of the pupae plus that of the infesting mite should not the direct effect on hygroregulation, the disruption of the honey- change much over time as the haemolymph would be transferred bee cuticular hydrocarbons, that was noted in this and other stud- from one to another. Therefore, on the ground of these results, ies, could have several side effects. In fact, the cuticle represents we cannot take for granted the hypothesis that haemolymph sub- the very first barrier for infecting pathogens, and the hydrocarbons, traction accounts on its own for the weight reduction observed in together with other compounds that are being identified (Turillazzi infested bees. Instead, since cuticle permeability is one major et al., 2006), may play a role. Under this respect, it is likely that not 1554 D. Annoscia et al. / Journal of Insect Physiology 58 (2012) 1548–1555

Fig. 9. Discriminant analysis of 15 individual honeybees differing for the infestation level during the pupal stage (uninfested bees: L; bees infested by one mite: L + 1 V; bees infested by three mites: L + 3 V). The groups are encircled arbitrarily. Groups are well separated suggesting an effect of mite infestation on cuticular composition.

4. Conclusions

In conclusion, our results show that the feeding activity of the mite, affecting the surface hydrocarbons, can alter the perfor- mances of the bee cuticle as regards as water stress tolerance; this, in turn, may affect the water content of the honeybee and its survival. According to our regression analysis of the relationship between weight at the eclosion and survival of infested bees, only a portion of the reduction in the longevity can be explained through the mechanism we described; further studies are therefore needed to accomplish a full comprehension of the proximate causes of bee mortality induced by the parasitic mite. V. destructor is, by far, the most important parasite of the honeybee and the dramatic consequences of Varroa infestation at colony level have already been described. Unfortunately, at the Fig. 10. Bee mortality during the ﬁrst week of life of uninfested (L) and infested (L + 1 V) bees that were maintained under different conditions of humidity (60% or moment, no control method can guarantee a satisfactory solution 75% R.H.). The effect of a lower rearing humidity was stronger in infested bees to this emergency. This work represents a preliminary step in the compared to uninfested ones. direction of an alternative solution to the Varroa problem, where the focus is on the elucidation of the negative effects of the parasite in view of developing possible strategies to mitigate the adverse only the water content of the honeybee is affected by the infesting outcome of parasitism. mite through the cuticle alteration but also its capacity to defend from pathogenic agents that are not transmitted through the food or by blood sucking parasites. Furthermore, surface hydrocarbons Acknowledgments are involved in the process of kin-recognition (Page et al., 1991; Dani et al., 2005) and task group differentiation (Kather et al., We gratefully acknowledge Dr. Rita Cervo for the valuable sug- 2011) that plays a fundamental role in the honeybee society. In this gestions and for critically reading the manuscript. This work was case, hydrocarbons were affected at a very early stage of the bee’s funded by the Italian Ministry of Agriculture (MIPAAF), (research life and this may not inﬂuence the subsequent acquisition of the project "Apenet - Ricerca e Monitoraggio in Apicoltura"). colony odor; nevertheless, any alteration of the chemical footprint of individual honeybees should be considered in light of the possi- References ble consequences on the social interactions taking place inside the hive. Bagnères, A.G., Blomquist, G.J., 2010. Site of synthesis, mechanism of transport and It has been shown that dietary hydrocarbons can be incorpo- selective deposition of hydrocarbons. In: Blomquist, G.J., Bagnères, A.G. (Eds.), rated in cuticular lipids (Blomquist and Jackson, 1973) and a rela- Insect Hydrocarbons: Biology, Biochemistry, and Chemical Ecology. Cambridge University Press, Cambridge (UK), pp. 75–99. tionship between cuticle composition and diet has already been Blomquist, G.J., 2010. Biosynthesis of cuticular hydrocarbons. In: Blomquist, G.J., demonstrated for the honeybee (Francis et al., 1989). Therefore, a Bagnères, A.G. (Eds.), Insect Hydrocarbons: Biology, Biochemistry, and Chemical possible intervention aiming at mitigating the negative effects of Ecology. Cambridge University Press, Cambridge (UK), pp. 35–52. mite infestation using a supplementary diet appears both reason- Blomquist, G.J., Jackson, L.L., 1973. Incorporation of labelled dietary n-alkanes into cuticular lipids of the grasshopper Melanoplus sanguinipes. Journal of Insect able and promising and surely worth of further investigation. Physiology 19, 1639–1647. D. Annoscia et al. / Journal of Insect Physiology 58 (2012) 1548–1555 1555

Bowen-Walker, P.L., Gunn, A., 2001. The effect of the ectoparasitic mite, Varroa Nazzi, F., 1992. Morphometric analysis of honey bees from an area of racial destructor on adult worker honeybee (Apis mellifera) emergence weights, water, hybridization in northeastern Italy. Apidologie 23, 89–96. protein, carbohydrate, and lipid levels. Entomologia Experimentalis et Applicata Nazzi, F., Milani, N., 1994. A technique for reproduction of Varroa jacobsoni Oud. 101, 207–217. Under laboratory conditions. Apidologie 25, 579–584. Comparini, A., Biasiolo, A., 1991. Genetic discrimination of Italian bee, Apis mellifera Nazzi, F., Brown, S.P., Annoscia, D., Del Piccolo, F., Di Prisco, G., et al., 2012. ligustica versus Carniolan bee, Apis mellifera carnica by allozyme variability Synergistic parasite-pathogen interactions mediated by host immunity can analysis. Biochemical Systematics and Ecology 19, 189–194. drive the collapse of honeybee colonies. PLoS Pathogens 8, e1002735. Dainat, B., Evans, J.D., Chen, Y.P., Gauthier, L., Neumann, P., 2012. Dead or alive: Page, R.E., Metcalf, R.A., Metcalf, R.L., Erickson, E.H., Lampman, R.L., 1991. deformed wing virus and Varroa destructor reduce the life span of winter Extractable hydrocarbons and kin recognition in honeybee (Apis mellifera L.). honeybees. Applied and Environmental Microbiology 78, 981–987. Journal of Chemical Ecology 17, 745–756. Dani, F.R., Jones, G.R., Corsi, S., Beard, R., Pradella, D., Turillazzi, S., 2005. Nestmate Reyment, R.A., 1989. Compositional data analysis. Terra Nova 1, 29–34. recognition cues in the honey bee: differential importance of cuticular alkanes Ribière, M., Ball, B.V., Aubert, M., 2008. Natural history and geographic distribution and alkenes. Chemical Senses 30, 477–489. of honey bee viruses. In: Aubert, M. et al. (Eds.), Virology and the Honey Bee. de Miranda, J.R., Genersch, E., 2010. Deformed wing virus. Journal of Invertebrate European Communities, Luxembourg, pp. 15–84. Pathology 103, S48–S61. Sage, R.D., 1982. Wet and dry-weight estimates of insects and spiders based on Del Piccolo, F., Nazzi, F., Della, Vedova.G., Milani, N., 2010. Selection of Apis mellifera length. American Midland Naturalist 108, 407–411. workers by the parasitic mite Varroa destructor using host cuticular Salvy, M., Martin, C., Bagnères, A.G., Provost, E., Roux, M., et al., 2001. Modifications hydrocarbons. Parasitology 137, 967–973. of the cuticular hydrocarbon profile of Apis mellifera worker bees in the Francis, B.R., Blanton, W.E., Littlefield, J.L., Nunamaker, R.A., 1989. Hydrocarbons of presence of the ectoparasitic mite Varroa jacobsoni in brood cells. Parasitology the cuticle and hemolymph of the adult honey bee (Hymenoptera: Apidae). 122, 145–159. Annals of the Entomological Society of America 82, 486–494. Sammataro, D., Gerson, U., Needham, G., 2000. Parasitic mites of honey bees: life Gibbs, A.G., 1998. Water-proofing properties of cuticular lipids. American Zoologist history, implications, and impact. Annual Review of Entomology 45, 519–548. 38, 471–482. Tentcheva, D., Gauthier, L., Zappulla, N., Dainat, B., Cousserans, F., et al., 2004. Hadley, N.F., 1994. Water Relations of Terrestrial Arthropods. Academic Press, San Prevalence and seasonal variations of six bee viruses in Apis mellifera L. and Diego, California. Varroa destructor mite populations in France. Applied and Environmental Kather, R., Drijfhout, F.P., Martin, S.J., 2011. Task group differences in cuticular lipids Microbiology 70, 7185–7191. in the honey bee Apis mellifera. Journal of Chemical Ecology 37, 205–212. Turillazzi, S., Mastrobuoni, G., Dani, F.R., Moneti, G., Pieraccini, G., et al., 2006. Le Conte, Y., Ellis, M., Ritter, W., Spivak, M., 2010. Varroa mites and honey bee Dominulin A and B: two new antibacterial peptides identified on the cuticle and health: can Varroa explain part of the colony losses? Apidologie 41, 353–363. in the venom of the social paper wasp Polistes dominulus using MALDI-TOF, Martin, S.J., Ball, B.V., Carreck, N.L., 2010. Prevalence and persistence of deformed MALDI-TOF/TOF, and ESI-Ion trap. Journal of the American Society of Mass wing virus (DWV) in untreated or acaricide-treated Varroa destructor infested Spectrometry 17, 376–383. honey bee (Apis mellifera) colonies. Journal of Apicultural Research 49, 72–79. Yang, X., Cox-Foster, D., 2005. Impact of an ectoparasite on the immunity and Martin, S.J., Highfield, A.C., Brettell, L., Villalobos, E.M., Budge, G.E., et al., 2012. pathology of an invertebrate: evidence for host immunosuppression and viral Global honey bee viral landscape altered by a parasitic mite. Science 336, 1304– amplification. Proceedings of the National Academy of Sciences 102, 7470– 1306. 7475. Möckel, N., Gisder, S., Genersch, E., 2011. Horizontal transmission of deformed wing Yang, X., Cox-Foster, D., 2007. Effects of parasitization by Varroa destructor on virus: pathological consequences in adult bees (Apis mellifera) depend on the survivorship and physiological traits of Apis mellifera in correlation with viral transmission route. Journal of General Virology 92, 370–377. incidence and microbial challenge. Parasitology 134, 405–412.