DIET SUPPLEMENTATION HELPS HONEY BEE COLONIES IN COMBAT INFECTIONS BY ENHANCING THEIR HYGIENIC BEHAVIOUR

The hygienic behavior in honey bees is a complex polygenic trait that serves as a natural defense mechanism against bacterial and fungal brood diseases and Varroa destructor mites infesting brood cells. The aim of this study was to evaluate the effect of a dietary amino acids and vitamins supplement “BEEWELL AminoPlus” on hygienic behavior of Apis mellifera colonies combating microsporidial and viral infections. The experiment was performed during a one-year period on 40 colonies alloted to ﬁ ve groups: one supplemented and infected with Nosema ceranae and four viruses (Deformed wing virus - DWV, Acute bee paralysis virus - ABPV, Chronic bee paralysis virus - CBPV and Sacbrood virus – SBV), three not supplemented, but infected with N. ceranae and/ or viruses, and one negative control group. Beside the l isted pathogens, honey bee trypanosomatids were also monitored in all groups. The supplement “BEEWELL AminoPlus” induced a signi ﬁ cant and consistent increase of the hygienic behavior in spite of the negative effects of N. ceranae and viral infections. N. ceranae and viruses signi ﬁ cantly and consistently decreased hygienic behavior, but also threatened the survival of bee colonies. The tested supplement showed anti-Nosema effect, since the N. ceranae infection level signi ﬁ cantly and consistently declined only in the supplemented group. Among infected groups, only the supplemented one remained Lotmaria passim -free throughout the study. In conclusion, diet supplementation enhances hygienic behavior of honey bee colonies and helps them ﬁ ght the most common infections of honey bees.


INTRODUCTION
Not only is the honey bee (Apis mellifera) a producer of honey, royal jelly and other bee products but is also the most important insect pollinator of both wild fl ora and crop species.For this reason, the decline in honey bee abundance which began at the end of the 20th century severely infl uenced the losses in agriculture [1].Thus, it is not surprising that increasing efforts are being made to fi nd out all putative factors contributing to the decrease in bee populations [2,3].Environmental pollutants such as agrochemicals [4][5][6] and heavy metals [7,8], together with pathogens [9,10], inadequate fl oral resources [11,12] and incorrect beekeeping practices [2] have been accused of having contributed to the decline in the number of bees.Among pathogens, there are endoparasites capable of compromising bee health and contribute to colony mortality, like microsporidia [13][14][15][16][17] and trypanosomatids [18][19][20], but also viruses [21][22][23][24][25].However, they do not necessarily infl ict damage per se, but may synergistically infl uence bees and even lead to colony death [26][27][28][29][30].
Hygienic behavior of honey bees is one of the many strategies they apply to effectively fi ght pathogens and presents their collective reaction to the presence of diseased brood [reviewed in [31][32][33].It results in the capability of the workers to detect the diseased brood and remove the larvae or, later, to open the wax capping of the cells and remove the pupae [reviewed in 34].By hygienic behavior, considered a social immune response of honey bees, honey bees fi ght against American foulbrood, chalkbrood, as well as against V. destructor [31][32][33][35][36][37][38][39].Being highly expressed in A. cerana than in A. mellifera, hygienic behavior may be one of the reasons underlying better health of the former [40].
Although assessed as low to moderately heritable, with a heritability (h 2 ) ranging from 0.17 to 0.65 [41][42][43][44][45], hygienic behavior is one of the most common traits selected for breeding programs [45][46][47].However, a great number of candidate genes have been associated with hygienic behavior [48][49][50][51][52][53][54][55] reaching as much as 73 genes [54].Being such highly polygenic, this behavior would be expected to be susceptible to external infl uences, but the investigations are scarce and the fi ndings non-consistent.Bigio et al. [56] reported no great infl uence of environmental conditions on hygienic behavior, but a reverse relationship was reported between hygienic behavior expression and altitude [57], seasonal and environmental variations [58].Hygienic behavior was not correlated with the colony strength [59] and was not affected by Nosema ceranae infection [60].However, the agrochemical imidacloprid was reported to signifi cantly impair hygienic activities of worker bees [61].
Beekeeping practices (colony manipulations, inadequate feeding, chemical and nonchemical 'alternative' treatments, migrations etc.) often negatively affect bee health [reviewed in 2 and 62; 63,64] but little is known about the infl uence of such activities on honey bee hygienic behavior.Sucrose syrup availability or scarcity and brood manipulation did not signifi cantly change the behavior [56].Neither of alternative treatments against Varroa and/or Nosema parasites, e.g.sugar dusting [65], thymol [66], chitosan and peptidoglycan [60], threatened the bees' hygienic potential, and thymol even increased the uncapping and removal of dead brood [66].However, migratory beekeeping practice seems to have a negative infl uence on hygienic behavior [58].
Given that the ecosystem is globally rather devastated, which inevitably leads to obvious lack of high-quality diverse bee forage [11], to meet the needs of the honeybees it is often resorted to the appliance of dietary supplements [2].Having in mind the increasing use of sugar syrup [67,68] which provides only energy, it is important to also provide mineral, vitamin and protein feed components.Thus, various dietary supplements have been tested for the ability to improve bee health, colony strength, food reserves and productivity [reviewed in 69] including pollen-substitute diets (protein supplements) and protein/vitamin supplements [70][71][72][73][74][75][76][77][78][79].
Due to the scarcity of data on the impact of supplementary diet on hygienic behavior of honey bees, the aim of this study was to evaluate the effect of a dietary amino acid and vitamin complex on hygienic behavior of full-sized free-fl ying colonies and their combat with N. ceranae and honey bee-associated viruses.During the study, honey bee trypanosomatids were also monitored knowing their common coexiste nce with Nosema sp.microsporidia [80][81][82].

Honey bee colonies
The research began in autumn 2019, on 80 honey bee (Apis mellifera) colonies originated from queens selected for hygienic behavior and located on the Pester Plateau (43°16'14'' N, 19°59'35'' E), Sjenica municipality, Serbia.Colonies were regularly checked for b oth bee and brood pathology by a veterinary specialist and were without signs of any disease including varroosis and infections caused by honey bee-associated viruses.The absence of Varroa infestation was proven by all three methods (debris, brood and bee examinations) recommended in COLOSS BEEBOOK [83].Prior to wintering, the colonies were supplied with optimum content of natural food resources (ca.19 kg of meadow honey and two frames fi lled with bee bread on both sides) without addition of sugar and/or dietary supplements.
In spring 2020 (end of March), a detailed inspection of the colonies was conducted, 40 colonies were selected according to adult bee population, brood areas and food amounts and allotted to 5 groups with 8 colonies in each (Table 1).The activities undertaken after the formation of the groups are shown in Table 2. On March 24, 2020 (Time 0), hygienic behavior was tested by the so called "pin-killed" technique in accordance with the procedure described by Kefuss et al. [84] and modifi ed by Stanimirovic et al. [36].Briefl y, on the one frame per each hive, the diamond area of comb (5 x 6 cm) was marked and all pupae within that area were killed with a pin.The frame was returned to the hive and after 24 hours checked.If more than 95% of the pin-killed cells were cleaned, the colony was considered super-hygienic, if the effi ciency of pupae removal was between 90% and 95%, the colony was proclaimed hygienic, while non-hygienic colonies were those which cleaned less than 90% of the sacrifi ced brood.The syrup was freshly prepared using 300 g meadow honey and 700 g ground sugar, dissolved in 1 liter of water at room temperature and administered.All groups except the negative control (C-) were infected: E1-NoV4+Beewell and E2-NoV4 were given the syrup which contained freshly made macerate of live bees infected with N. ceranae and four viruses: (Deformed wing virus -DWV, Acute bee paralysis virus -ABPV, Chronic bee paralysis virus -CBPV and Sacbrood virus -SBV).Group E3-No was infected with N. ceranae only, and E4-V4 only with the four viruses.The negative control group (C-) was left uninfected but was given the water-sugar-honey syrup, without the addition of dietary supplements.
On day 21 post infection, on April 17, 2020 (Time 1) hives were tested for hygienic behavior using the pin-killed brood technique [36].On the same day, bee samples were taken for laboratory analyses and a dietary amino acid and vitamin supplement Beewell AminoPlus (Provet Genome Biotechnology Laboratory, Ankara, Turkey), was administered to Group E1-NoV4+Beewell, but not to other groups (Table 1).The supplement was applied to the same group two more times in the same quantities in 7-day intervals, according to the instructions of the producer.
In late summer, on August 25, 2020 (Time 2), the colonies were tested again for hygienic behavior, and bees were sampled for laboratory analyses.
From March 8, 2021 the syrup with supplement was administered again (three times in a 7-day interval).On day March 29, 2021 that is 7 days after the third application (Time 3), hygienic behavior was assessed and bees sampled for analyses.

Preparation of the inoculum
The inoculum for the artifi cial viral infection of the bees was prepared according to de Miranda et al. [85].From the hives where all the viruses were present (confi rmed by PCR) 200 bees (approximately 20 g) were collected from the hive entrance.The bees were macerated and the virus quantity was measured in the suspension.The bees were infected with a volume suffi cient to ensure that each bee received the minimum infective dose of 10 6 -10 11 particles [86,87].For Nosema infection, the suspension of N. ceranae spores was prepared as described in Fries et al. [88] and added into the syrup to obtain a fi nal concentration of 10 6 spores/ml.

Bee sampling for laboratory analyses for the presence of pathogens
Bees were sampled three times (Time 1, Time 2 and Time 3).Each time, approximately 100 live forager bees were sampled from each colony, directly from the hive entrance after closing it for 20-30 min [89].Live bees were collected in sterile single-use vessels, immediately stored in dry ice, transported to the laboratory and stored at -20°C until processed.

Detection of Nosema spores and determination of colony level infection
Abdomens of 60 bees from each colony were macerated in 5 ml of water and the suspension was examined microscopically at 400× magnifi cation.In cases of Nosemapositive samples, the colony level infection was determined by hemocytometer through the average number of spores per bee in a pooled sample obtained using 60 bees macerated in 60 ml of water (OIE, 2018).The suspensions of all samples were further used for DNA extraction and PCR analyses.

PCR detection and identifi cation of honey bee microsporidian and trypanosomatid parasites
DNA was extracted from 1 mL of sample suspension obtained in the previous step and using DNeasy Plant Mini Extraction Kit (Qiagen, Hilden, Germany) as in Stevanovic et al. [90,91].For confi rmation of Nosema ceranae species PCR-RFLP with nos-16S-fw/rv primers was applied as in Stevanovic et al. [91], while for the detection of Lotmaria passim or Crithidia mellifi cae, PCR protocols with primer pairs CmCytb_F/R and LpCytb_F1/R respectively were used as described in Stevanovic et al. [80].All PCR amplifi cations were performed in T100™ Thermal Cycler (Bio-Rad, Germany).

RT-PCR detection and identifi cation of honey bee viruses
From each sample, 30 randomly selected bees were crushed and homogenized in a sterile mortar in the presence of 5 ml PBS solution.After homogenisation and centrifugation for 15 min at 5,000×g, 140 μl of supernatant was collected and used for RNA extraction.Total RNA was extracted using ZR Viral RNA Kit™ (Zymo Research, Orange, CA).The average of 2 μg of extracted RNA was used for a single real-time RT-PCR reaction.
Thermal amplifi cations were performed in Rotor-Gene Q 5plex (Qiagen, Germany) and the presence of DWV, ABPV, CBPV and SBV in bee samples were tested using the Rotor-Gene Probe RT-PCR Kit (Qiagen, Germany), in separate single-step reactions.
The primer pairs, probes and thermal protocols were as in our previous work of Cirkovic et al. [92].

Statistical methods
The results for N. ceranae infection level (spore counts) and hygienic behavior expression level were tested for normality by using Shapiro-Wilk's test.Given that data for N. ceranae spore counts were not normally distributed (Shapiro-Wilk's test, p<0.05), log10 transformation was applied, and groups were compared in two-way ANOVA with repeated measures in one factor, followed by Tukey's test.Data for the behavior were compared between the groups over time using one-way ANOVA followed by Tukey's test, and within the group over time using one-way repeated measures ANOVA followed by Tukey's test.Fisher's exact test was used to compare differences in the occurrence of honey bee viruses and L. passim between groups.The levels of signifi cance below 0.05 (p<0.05) were considered signifi cant.Statistical analysis of the results obtained in the experiment was carried out using statistical software GraphPad Prism version 6 (GraphPad, San Diego, CA, USA).
Ethical approval: The research has been conducted on invertebrates and in compliance with all the relevant national regulations and institutional policies.

Hygienic behavior
The level of expression of hygienic behavior differed signifi cantly (p<0.05) between all assessment times (Time 0, Time 1, Time 2 and Time 3) within each group except for the negative control (C-).In all groups that were artifi cially infected, but not supplemented (E2-NoV4, E3-No and E4-V4) the level of expression of hygienic behavior consistently decreased through the whole experimental period and was signifi cantly (p<0.05)lower in each subsequent assessment time compared to the previous one (Table 3, Figure 1).In group E2-NoV4, all colonies died before the last assessment time, so it was not possible to assess the behavior in Time 3. However, in the supplemented group (E1-NoV4+Beewell), the level of expression of hygienic behavior decreased only before supplement application (from 97.52±0.77 in Time 0 to 87.40±0.86 in Time 1), and after the application of supplement, the behavior consistently increased, being signifi cantly greater in Time 2 (91.11±1.38)compared to Time 1, and in Time 3 (92.98±1.65)compared to both Time 2 and Time 1 (Table 3, Figure 1).At the Time 0, hygienic behavior signifi cantly differed (p<0.05)only between E3-No and C-groups.At the Time 1, Time 2 and Time 3, hygienic behavior was most expressed in C-group (97.31±0.60,97.21±0.43 and 96.38±0.43,respectively) and was signifi cantly (p<0.05)higher in comparison to all other groups (Table 3, Figure 2).Among the infected groups, hygienic behavior at Time 1 was most expressed in groups E3-No (88.64±1.30)and E1-NoV4+Beewell (87.40±0.86)without signifi cant (p>0.05)difference between them, but at Time 2 and Time 3, the group E1-NoV4+Beewell took the lead, reaching the highest level of hygienic behavior (91.11±1.38 and 92.98±1.65,respectively) that was signifi cantly (p<0.05)higher compared to all other infected, but not supplemented groups (Table 3, Figure 2).

Nosema ceranae infection
In all groups, N. ceranae infection level (log 10 ) differed signifi cantly (p<0.05) between all sampling times (Time 1, Time 2 and Time 3) as shown in Table 4 and Figure 3.Only in supplemented group (E1-NoV4+Beewell), N. cerane infection level consistently decreased during the experiment, as was the highest at Time 1, lower at Time 2 and the lowest at Time 3. In non-supplemented groups, either infected (E2-NoV4, E3-No, E4-V4) or not (C-), the situation was reversed: infection level raised through time, reaching the highest value at the last sampling time (Time 3).
was signifi cantly (p<0.05)lower than in groups artifi cially infected with N. ceranae and not supplemented (E2-NoV4 and E3-No).At Time 3, the lowest infection level was recorded in E1-NoV4+Beewell group and that level was signifi cantly (p<0.05)lower than in all other groups, including the negative control (E2-NoV4, E3-No, E4-V4 and C-).The highest Nosema-infection level in all sampling times was in groups that were artifi cially infected with N. ceranae but were not supplemented (E2-NoV4 and E3-No); the infection level in those groups was signifi cantly (p<0.05)higher compared to both, C-and E4-V4 groups.Between groups E2-NoV4 and E3-No there were no signifi cant differences (p>0.05) in Nosema-infection level in any of sampling times (Table 4, Figure 4).Group names are indicated in Table 1.

Viruses
DWV: In all groups except in the negative control (C-), there were no signifi cant (p>0.05)differences in the presence of DWV through time, i.e. when the results were compared between sampling times (Time 1, Time 2 and Time 3).In contrast, signifi cant (p<0.05)differences were revealed within C-group when DWV presence was compared between Time 2 and Time 1 and between Time 3 and Time 1 (Figure 5).
ABPV: Looking at the presence of ABPV through the time (Figure 5), i.e. between sampling times (Time 1, Time 2 and Time 3), only in E1-NoV4+Beewell group there was a consistent decrease in the percentage of ABPV-positive hives and the difference between Time 1 (100% infected hives) and Time 3 (12.50%infected hives) was signifi cant (p<0.05).In all other groups, there were no signifi cant (p>0.05)differences in percentage of ABPV-positive hives between sampling times.
In each sampling time there were signifi cant differences (p<0.05) in the presence of ABPV (Figure 5): at Time 1, between 100% ABPV-positive groups (E1-NoV4+Beewell, E2-NoV4 and E4-V4) and viruses-free groups (E3-No i C-); at Time 2 the situation changed only in supplemented group (E1-NoV4+Beewell) in a sense of decreasing Group names are indicated in Table 1.
the ABPV presence to 50% hives; fi nally, at Time 3 all fi ve groups were infected, but only E2-NoV4 and E4-V4 remained 100% infected with ABPV, in E1-NoV4+Beewell the percentage of positive hives decreased, and in groups E3-No and C-, the ABPV virus appeared for the fi rst time (12.5% and 37.50% positive hives, respectively).The difference between 100% infected group and any other group was signifi cant (p<0.05).
CBPV: Comparisons of the CBPV presence in different sampling times (Time 1, Time 2 and Time 3) revealed signifi cant differences only in cases of groups E3-No and C-(Figure 5).In fact, group E3-No was CBPV-free at Time 1, but later became infected Group names are indicated in Table 1.
(100% and 87.50% infected hives in Time 2 and Time 3, respectively) and signifi cantly (p<0.05)different compared to Time 1.In the C-group, infection with CBPV was registered only at Time 3 (100% infected hives) which was signifi cantly (p<0.05)higher compared to Time 1 and Time 2. It is worth to emphasize a consistent decline of CBPV presence in the supplemented group E1-NoV4+Beewell, from 100% infected hives (at Time 1) to 87.50% (at Time 2) and fi nally to 62.50% (at Time 3).
At Time 1 and Time 2, there were signifi cant (p<0.05)differences between CBPVinfected groups and CBPV-free groups (Figure 5).
SBV: All hives in E1-NoV4+Beewell group were SBV-positive at Time 1; 50% of hives remained SBV-infected at Time 2, but only 12.50% at Time 3. The decrease at Time 3 was signifi cant (p<0.05)compared to Time 1.In E3-No group, that initially (at Time 1) was viruses-free, 50% of the hives appeared SBV-infected at Time 2 and 62.50% of the hives at Time 3; the last percentage (62.50%) is signifi cantly (p<0.05)higher than initial 0% (Figure 5).
Signifi cant (p<0.05)differences were affi rmed between SBV-positive groups and SBVfree group at Time 1 and Time 2, and also at Time 3 when groups E2-NoV4 and E4-V4 were compared with C-group (that was SBV-free) and supplemented group E1-NoV4+Beewell that contained 12.50% hives infected with SBV (Figure 5).

Trypanosomatids
Only L. passim was detected and not C. mellifi cae.Signifi cant (p>0.05)differences in presence of L. passim were affi rmed only in groups E2-NoV4 and E3-No between Time 1 and Time 3 and between Time 2 and Time 3 (Figure 5).

DISCUSSION
Hygienic behavior in honey bees is a complex, disease-resistant, polygenic trait [45,48,49] which genetic basis was investigated by many genomic and tanscriptomic studies [48][49][50][51][52][53][54][55].The greatest number of candidate genes suggested to contribute to hygienic behavior is 73 as revealed by high-depth full-genome sequencing in a study of Harpur et al. [54].Moreover, Guarna et al. [45] discovered robust protein expression markers as a completely new tool to select for this behavioral trait.
In this study, we investigated the potential of diet supplementation to enhance the hygienic behavior of honey bee colonies and thus help them fi ght the most common infections.There were fi ve groups in the experiment: the group that received supplement (E1-NoV4+Beewell) was artifi cially infected with N. ceranae and four viruses (DWV, ABPV, CBPV and SBV); three groups were not supplemented, but were artifi cially infected with N. ceranae and viruses (E2-NoV4), only with N. ceranae (E3-No) or only with viruses (E4-V4); negative control group (C-) was neither supplemented, nor infected.
The results indicate that the tested supplement Beewell AminoPlus signifi cantly stimulates hygienic behaviour.In fact, starting from the day of the supplement application (on 17 April 2020 -Time 1) in E1-NoV4+Beewell group, the hygienic behaviour was signifi cantly better expressed in each subsequent assessment time (Time 2, Time 3) compared to previous one(s), i.e.Time 1, Time 1 and Time 2, respectively.In contrast, in all other (not supplemented) groups (E2-NoV4, E3-No, E4-V4 and C-) the behavior consistently decreased; in the infected groups E2-NoV4, E3-No, E4-V4, it was signifi cantly lower in each subsequent assessment time compared to each previous time point (Figure 1).Hygienic behavior decline in non-supplemented infected groups was much more intensive than its increase in the supplemented group (Table 3, Figure 1) suggesting that it is easier to worsen than to improve this behavior.Nevertheless, only imidacloprid has been reported to signifi cantly impair hygienic activities of worker bees [61], and further studies should investigate the infl uence of other external factors, both environmental and beekeeping-induced, on hygienic behavior.
The benefi cial impact of the Beewell AminoPlus on the hygienic behavior is evident from the results recorded for Time 2 and Time 3 (Figure 2) that indicate the highest level of behavior in the supplemented group and signifi cantly greater that in other infected, but not-supplemented groups.Thus, the supplement stimulated hygienic behaviour of in spite of the negative infl uence of infections.However, super-hygienic level (>95%) remained only in the negative control group during the entire experiment, while supplemented group reached "hygienic" level (90-95%) in Time 2 and Time 3 but not re-achieved the super-hygienic level.
Nevertheless, this achievement of Beewell AminoPlus is better than other tested supplements.In fact, thymol showed the potential to improve the uncapping and removal of freeze-killed brood but 88% was a maximal removal rate achieved [66], whilst chitosan and peptidoglycan did not alter the hygienic behavior of bees infected with Nosema [60].
Changes of the hygienic behavior expression level (consistent decrease) and N. ceranae infection level (consistent increase) through time (Figures 1 and 3, respectively) indicate a suppressive effect of infections (induced either by N. ceranae only, or by viruses only or by mixed infection of N. ceranae and viruses) on hygienic behavior.In contrast, N. ceranae infection did not affect hygienic behaviour in study of Valizadeh et al. [60].This disagreement is probably caused by completely different method applied for behaviour assessment.The most important difference is the assessment time frame: in our study bees were evaluated for hygienic behaviour: 21 day, four months and almost a year following artifi cial infection (Time 1, Time 2 and Time 3, respectively), while in the study of Valizadeh et al. [60] bees were only 12-15 days old and even the authors have questioned the impact of N. ceranae on bee health in such short period of infection.
In our study, the strongest negative effect of N. ceranae and viral infections was recorded in group E2-NoV4 (with mixed artifi cial infections of N. ceranae and four viruses) since all colonies died until the last time point (Time 3).In that group, beside N. ceranae and viral infections that were artifi cially induced, L. passim was confi rmed, so it could also contribute to the mortality.
Tested supplement showed benefi tial effect in terms of control of N. ceranae infection; in supplemented group (E1-NoV4+Beewell) Nosema spore load signifi cantly and consistently declined through the time, while in all other, non-supplemented groups (E2-NoV4, E3-No, E4-V4 and C-), the situation was inverted (Figure 3).The look on the Nosema level data in E1-NoV4+Beewell group in different time points (Figures 3) reveals a great Nosema-control effect of applied supplement in wide time-frame after the application, since signifi cant decrease of N. ceranae load is evident both after fi ve months of fi rst application set and after 7 days of second application set.Besides, comparison within groups in each time point (Figure 4) revealed that in Time 3 convincingly the lowest N. ceranae load was in the supplemented group.Beewell AminoPlus has been already reported to reduce N. ceranae spore number in cage experiment, but also to prevent Nosema-induced host immunosuppression by modifying the expression of immune-related genes (those that code abaecin, hymenoptaecin, defensin, apidaecin and vitellogenin) in Nosema-infected bees [70].We might assume that in current study the supplement Beewell AminoPlus also potentiated Nosema control through enhancing individual bee immunity by up-regulation immune-related genes.Beside those genes that are quite proven to be down-regulated by Nosemaparasitism [5,6,30,71,72,[93][94][95][96] other host genes are also prone to be suppressed by the same factor: genes involved in chitin metabolism and cuticle coatings [95,96], genes related to metabolism of carbohydrates [95,96]; genes encoding odorant binding proteins [96] and genes involved in homeostasis of intestinal tissue, cell apoptosis and renewal [97][98][99].We may hypothesize that in the current study Beewell AminoPlus achieved positive effects by stimulation of some of these genes, but further studies are necessary to investigate that.It would be interesting to assess behavioural regulation of the intake of supplement as in de Sousa [100], and the potential on health and reproductive parameters generally affected by supplements [101,102].
Regarding the viruses (Figure 5), it seems that the supplement did not affect DWV, but contributed to the consistent decrease of ABPV, CBPV and SBV (signifi cant in cases of ABPV and SBV presence between Time 1 and Time 3).L. passim appeared in all experimentally infected groups, except the supplemented one (Figure 5).In E2-NoV4 and E3-No groups L. passim infection was recorded as early as in Time 2 in 37.50% of hives and reaching its presence in 100% of hives in Time 3 that was a signifi cant increase compared to Time 2. In E4-V4 group, L. passim was confi rmed only at the fi nal time point (Time 3) in 50% of hives.In relation to the absence of L. passim in E1-NoV4+Beewell group, we may assume that the supplement was the one that prevented the invasion of that parasite.

CONCLUSION
The impact of diet supplementation on hygienic behavior of honey bee colonies has been poorly known.The supplement tested in this study expressed a positive infl uence on hygienic behaviour of colonies infected with N. ceranae and four viruses.
In fact, starting from the day of the supplement application, the hygienic behaviour was signifi cantly better expressed in each subsequent assessment time compared to previous one(s), contrary to non-supplemented groups, either infected or not.The supplement also helped colonies in terms of control of N. ceranae infection.Further studies are needed to investigate if the supplement has potential to prevent L. passim infection.

Table 1 .
Description of groups

Table 2 .
Experimental design

Table 3 .
Comparison of hygienic behavior between assessment times and between groups

Table 4 .
Comparison of N. ceranae infection level (log 10 ) between sampling times and between groups A, B, C Different superscript capital letters indicate significant differences p<0.05 between sampling times within each group.a,b,c, d Different superscript lowercase letters indicate significant differences p<0.05 between groups within each sampling time.Group names are indicated in Table1.