HONEYBEE VIRUSES PRESENCE IN SERBIAN APIARIES: A REVIEW

Th e honey bee Apis mellifera is an important benefi cial insect recognized by production of honeybee products, having an important role in agricultural production as a pollinator, and playing an equally crucial role in conserving the biodiversity in many natural ecosystems. According to available literature data, during the period of more than 30 years, dramatic losses of honey bee winter colonies have frequently been reported all over the world, which could have a direct infl uence on human food resources and can aff ect not only apiculture or agriculture, but also pose an anthropological threat. One among many reasons for global bees-decline phenomenon is the infl uence of many viruses on honeybees’ health. Until today, at least 24, and even more viruses were detected in honeybees, and for many of them the pathogenicity and impact on honeybees’ health still remain unknown. However, it is well known that some of these viruses like acute bee paralysis virus, chronic bee paralysis virus, deformed wing virus, black queen cell virus, sacbrood virus, Kashmir bee virus, Israeli acute paralysis virus, slow bee paralysis virus, Varroa destructor virus-1 and some others have direct or indirect infl uence on individual honeybee or on whole honeybees’ colony health. In this paper, an overview of existing literature data on the presence, prevalence and characterization of honeybee viruses detected in honeybee colonies and apiaries from diff erent regions in Serbia from fi rst detection of their presence in 1986 till nowadays is presented and discussed.


INTRODUCTION
Th e western honey bee Apis mellifera is an important benefi cial insect recognized by production of honeybee products, which plays an essential role in the ecology of natural environments and in agricultural production through pollination (Petrović et al., 2013;Ćirković et al., 2018;Chapman et al., 2019). Almost one-third of distinctive agricultural crops need bees' pollination (O'Neal et al., 2018). Pollination services are mandatory for the production of crops like fruits, nuts and fi bres, but also the production of many other agricultural crops is signifi cantly improved by pollination (Tantillo et al., 2015). Also, honeybees play an equally crucial role in conserving the biodiversity in many natural ecosystems. Th e value of bee pollination services is commonly measured in billions of dollars representing about 9.5% of the value of crops across the world (Tantillo et al., 2015). In United States, the value of honeybee pollination for agriculture has been estimated at more than 14.6 billion dollars (Morse and Calderone, 2000). According to FAO and the European Union, the value of pollination is 20 -30 times higher than the value of honey production (Antúnez et al., 2012).
According to available literature data, in the last 30 years, dramatic honey bee winter colony losses have been reported frequently from diff erent regions all over the world (Breeze et al., 2014). Pollinating insect declines have a direct infl uence on human food resources and can aff ect not only apiculture or agriculture but also has anthropological threats (Francis et al., 2013; McMenamin and Flenniken, 2018). Among reasons for global bees-decline phenomenon several factors could be appointed like pesticides, destruction of habitat, modern industries, intensive agriculture, parasites/pathogens, climate change, and inadequate food supply (Ullah et al., 2021). Th e honeybees' health is often compromised by diff erent pathogens, such as the mites Varroa destructor and Acarapis woodi, the microsporidia Nosema ceranae and Nosema apis, the bacteria Paenibacillus larvae and Melissococcus plutonius and diff erent viruses (Antúnez et al., 2012). Many researchers demonstrated that viral infections in honey bee colonies are considered a key risk for their health at both individual and colony level (Tantillo et  Viruses are mostly the hidden enemies of honey bees as compared to other pathogens, because most of infections pass without clinical manifestation or characteristic disease signs (Ullah et al., 2021). During the last two decades, honeybee virus infections have been increasingly investigated and have emerged as one of several causes of the honeybee colony losses (Tantillo et al., 2015). However, viruses are probably the least understood part of honeybee pathology mainly because of the lack of information of the objective data about viral disease outbreaks (Toplak et al., 2012). During the last few years, our understanding of the diversity of viruses infecting bee species has grown dramatically, especially due to rapid improvements and increasing accessibil-ity of next-generation sequencing approaches. So far, honey bees have been reported to be the host to at least 24 and even more (Beaurepaire et al., 2020) viruses, primarily positive-strand RNA viruses belonging to order Picornavirales, and in the families Dicistroviridae and Ifl aviridae. Th ese ones are able to infect the diff erent developing stages of the honeybees, including eggs, larvae, pupae and adults (Allen and Ball, 1996 (Grozinger and Flenniken, 2019). Many of these viruses are also found in plants and fungi, and further studies will be needed to describe the role of many newly detected viruses and their potential infl uence on bees' health.
Oft en, clinically visible symptoms of honey bee virus diseases are mostly associated with other infectious agents, such as the presence of microsporidia Nosema apis and strong infestations with Varroa destructor mites (Ullah et al., 2021). Several studies implicate that the combination of certain virus infections and Varroa destructor infestation represent a serious threat to honeybee health (Allen and Ball, 1996 ). An established opinion among researchers is that Varroa in association with a range of hon-eybee viruses is a signifi cant factor in the losses of managed honeybee colonies seen globally. Th e spread of this mite to the Western honeybee -Apis mellifera, and its ability to act as a viral reservoir, incubator, activator and transmitter has resulted in levels of certain viruses that aff ect the survival of the colony (Tantillo et al., 2015). . In addition to this direct transmission routes, some viruses, like DWV, mainly spread among honeybees via Varroa destructor mites that become infected from infected bees and behave as a carrier to spread viruses to healthy bees when they feed on them (vector-borne transmission) that represents an example of indirect transmission (Ullah et al., 2021).
In nature, BQCV, DWV, KBV, and SBV infect larvae and pupae as well as adult bees, while ABPV aff ects only adult bees (Chen et al., 2004). Th e fi rst detected honeybee virus was SBV detected at the beginning of 20th century (1913) in USA (White, 1913), much earlier than many of human and other animal viruses. SBV aff ects larvae of honeybees and causes sacbrood disease. Aff ected larvae change from pearly white to grey and fi nally black. When affected larvae are carefully removed from their cells, they appear to be a sac fi lled with water. Th e adult bees develop a latent infection characterized only by a decreased life span, without acclaimed symptoms (Berènyi et al., 2006). Th is latent infection is very important for the transmission of SBV since virus is accumulated in the head and in the hypopharyngeal glands of infected nurse bees that are responsible for feeding the larvae (Tantillo et al., 2015). Also, a positive correlation was observed between the intensity of V. destructor prevalence and the presence of SBV in adult bee samples (Tentcheva et al., 2004). Th e frequencies of SBV infection were much higher during the spring period, when the brood season begins and large numbers of susceptible larvae and young adults are present (Berènyi et al., 2006) BQCV was fi rst detected in brown and black coloured queen larvae and prepupae but it also aff ects larvae and pupae of worker bees without causing signs (Antúnez et al., 2012). In honeybee colonies, BQCV is more prevalent in adult bees than in the pupae, although it clinically aff ects mainly developing queen larvae and pupae of the queen, representing the primary hosts of the virus. Th ese larvae acquire a pale yellow appearance and the symptoms are similar to those caused by SBV infection (Tentcheva et al., 2004;Tantillo et al., 2015). An important correlation was observed between the incidence of BQCV and Nosema apis in honeybee colonies with a peak of infection and infestation during spring and early summer (Allen and Ball, 1996;Tantillo et al., 2015).
DWV was fi rst isolated in Japanese apiaries from adult honeybees with a particular deformity of wings (Bailey and Ball, 1991), and is one of the most widely distributed honeybee viruses around the globe (Tantillo et al., 2015). Th e presence and high prevalence of DWV is reported in all continents except in Oceania (Allen and Ball, 1996). DWV has been detected in more than 20 bee species, and there is evidence that it replicates in several of these species and causes damaging symptoms in two bumble bee species (i.e., Bombus terrestris and Bombus pascuorum), as well as in A. mellifera (Grozinger and Flenniken, 2019). Infection with DWV is typically associated with the presence of V. destructor, and results in wings deformation in honeybees. Also, it is one of the most studied viruses that aff ect honeybees due to its relation with colony losses induced by honeybee's mite V. destructor (Antúnez et al., 2012). Th ere are several routes of DWV transmission that can infl uence its abundance and virulence. In bee colonies, DWV is transmitted vertically from queen or drone to the off spring, and horizontally via trophallaxis and shared food resources, but very oft en by Varroa destructor as their biological vector (Grozinger and Flenniken, 2019). DWV is able to infect all bee developmental stages from eggs to adults, even if it shows a higher replication in pupae. DWV is a virus with a low pathogenicity and is oft en responsible for latent infections that can appear in clinical form aft er a stressful situation such as high infestation with Varroa destructor and some others (Tantillo et al., 2015). DWV levels in bees parasitized by Varroa are signifi cantly higher than levels of bees infected by other routes. Also, Varroa appear to benefi t from this interaction because they produce more off spring while feeding on DWV-infected pupae (Grozinger and Flenniken, 2019) CBPV was fi rst isolated in 1963 from diseased honeybees (Bailey et al., 1963) and it is present at all continents. Th is virus can persist throughout the years as a subclinical infection, and the prevalence of the infection did not follow any seasonal pattern (Tentcheva et al., 2004;Tantillo et al., 2015). At low levels, infected colonies do not show clinical signs, but high level of those viruses induces high mortality rates. CBPV induces abnormal trembling of the wings and body of honeybees and sometimes black individuals crawling at the hive entrance can be detected. Th ese are the main two clinical symptoms in bees infected with CBPV. Th e wings are partially spread or dislocated. In some cases, the crawling bees can be in large numbers representing the clinically evident disease (Antúnez et al., 2012). CBPV was detected in queens and all their off spring at all developmental stages including eggs. Crowded condition of the colonies promotes the spread of the virus by direct contact of healthy bees with paralysed individuals, but vertical virus transmission is also possible. Interestingly, CBPV infections have never been related to Varroa destructor infestations and the virus has not been reported (Tantillo et al., 2015).
ABPV is a common infective agent of honeybees, frequently detected in apparently healthy colonies ( IAPV was fi rst isolated in 2004 from Israeli apiaries, where it caused a signifi cant mortality in honeybees (Tantillo et al., 2015). Th e virus was found to be present in every developmental stage of honey bee and its infection was observed almost in all tissues, but mostly existed in hypopharyngeal glands, alimentary canal and nervous system (de Miranda et al., 2012). IBPV is closely related to KBV and ABPV. Besides Israel, IBPV is widespread in Australia and USA. Like ABPV and many other bee viruses, IBPV usually persists in bee colonies at low titres without clinical symptoms. However, diff erent stress factors infl uencing the weakening of the honeybee defences can exacerbated the IBPV infection that leads to death of the honeybees. Clinical symptoms of such infection are very similar to the ABPV infection with rapidly progressing paralysis, including trembling, inability to fl y, gradual darkening and loss of hair of the thorax and abdomen resulting in massive deaths of adult bees (Tantillo et al., 2015).
KBV is endemically present in Australia and in the USA, and has been reported in Europe only rarely (Tentcheva et al., 2004;Berènyi et al., 2006). Like IBPV and ABPV, KBV persists at low titres in apparently healthy colonies until diff erent stress factors activate the viral multiplication causing the clinical manifestation of disease and death of the colony. Diff erent developing stages of bees could be hit by infection, but without clearly defi ned disease symptoms. Th e transmission of KBV in naturally infected colonies can occur via multiple routes including foodborne transmission, but also transmission by Varroa destructor as physical or true vector of this virus. Recently, KBV was proved to be highly important marker of Colony Collapse Disorder (CCD), i.e., rapid loss of the colony's adult bee population (Tantillo et al., 2015).
Slow bee paralysis virus (SBPV) primarily aff ects the forelegs (paralyses) of honey bees but can also be found in the head, salivary gland, mandibular and hypopharyngeal glands, crop, fat body, while present in the thorax, midgut, hindlegs and rectum in low quantity. It was correlated with colony collapse in England, but it is mostly less prevalent in other European apiaries (de Miranda et al., 2012).
Th e diagnosis of bee virus infections is diffi cult because honey bee viruses usually persist as unapparent infections and cause no clinical signs of disease, and because the fact that bee colonies can be simultaneously attacked by more Th e increased losses of bee colonies have been also reported in Serbia during the last two decades. Because of the eff ects of viral infections on honeybee health and honeybee colonies losses, the aim of our study was to assess the presence and distribution of diff erent honeybee viruses in Serbian apiaries. In this paper, the authors presented an overview of existing literature data on the presence, prevalence and characterization of honeybee viruses detected in honeybee colonies and apiaries from diff erent regions in Serbia, from the fi rst detection of their presence in 1986 till nowadays.

THE PRESENCE OF HONEY BEE VIRUSES IN SERBIAN APIARIES
By reviewing the available literature data only limited number of studies on investigating the presence of viruses in Serbian honeybee colonies and apiaries were found. Th e fi rst detection of viruses in Serbian apiaries has been done by Kulinčević and co-workers even 35 years ago (Kulinčević et al., 1990).
In that study, samples of adult honeybees were collected from 2 apiaries near Belgrade in autumn 1986 and early spring 1987, and tested on virus presence by electron microscope and immunodiff usion test (with diff erent bee virus antisera). Th e results of the study confi rmed the presence of 4 bee viruses, being: ABPV, Cloudy Wing Virus (CWV), DWV (at that time known as J strain of Egypt bee virus) and BQCV in Serbian apiaries.
Aft er a long time period without any published data on the presence of honey bee viruses in Serbia, the fi rst surveillance study on viruses presence based on molecular diagnostic methods was done in apiaries in two districts in Vojvodina Province, northern part of Serbia (Petrović et al., 2013). In this study, 30 bee samples originating from 15 apiaries at 13 locations in districts of South Bačka (7) and Srem (6)  locations on the territory of Srem District were slightly diff erent. Th e most prevalent viruses were BQCV in 100% (10/10) and DWV in 90% (9/10) tested samples, followed by ABPV found in 70% (7/10), SBV found in 40% (4/10) and CBPV found in only 1 out of 10 (10%) examined samples. Besides BQCV, the most prevalent bee virus found in South Bačka District was CBPV and in Srem District it was DWV. Th e authors also pointed out that this diff erence could not be seen in clinical manifestation of the disease on the fi eld. Th e lack of differences in clinical manifestations might be attributed to the relatively small number of examined samples or insuffi cient anamnestic data obtained from the fi eld. If such diff erences really exist and remained unnoticed, it might be infl uenced by some factors that were not identifi ed at that time. Further on, the data obtained in this study showed that in most of the tested samples (96.67%), more than one virus was identifi ed. Only one analysed sample (3.4%) was infected with only one virus, 2 (6.7%) with two viruses, 14 (46.7%) with three viruses, 6 (20%) with four viruses, and 7 (23.3%) samples were found positive on fi ve viruses simultaneously. Almost the half of the analysed samples had 3, and more than 40% of the analysed samples had 4 or 5 simultaneous viral infections. Th e severity of clinical manifestations with high bee losses was associated with higher amount of viruses detected in the samples (Petrović et al., 2013). Th e results obtained in this study indicated high prevalence of 5 out of 6 examined bee viruses in Serbian apiaries, and the great burden . Contrary to these fi ndings, ABPV was not detected in any of the samples from the western most sampled location (Prijepolje). Moreover, simultaneous infections with DWV and ABPV were identifi ed in 50.9% of samples (28/55 colonies). Single infections with DWV were found in 23.6% samples (13/55 colonies) and with ABPV in 9.1% samples (5/55 colonies). At least one virus was detected in 85.5% samples and in 85.45% (47/55) colonies. Only 14.5% (8/55) of sampled colonies were negative for the presence of tested viruses (Simeunović et al., 2014). Th e authors concluded that high incidence of DWV and ABPV positive samples found in clinically asymptomatic colonies can be the consequence of ineffi cient and delayed Varroa treatment since the role of this mite in the transmission and activation of honey bee viruses is well known.
Similar, yet more extensive research was done by Ćirkovic et al. (2018), who studied the prevalence DWV, CBPV, ABPV and SBV in colonies of different strength located in fi ve regions of Serbia. In addition, the genetic relationship between nucleotide sequences of detected bee virus strains from Serbia and those published in NCBI GenBank from diff erent parts of the world was investigated. In total, 150 colonies from 32 apiaries (about 5 colonies per apiary) located in 5 administrative regions in Serbia (Vojvodina Province as Northern Serbia, Western Serbia, Central Serbia, Eastern Serbia and Southern Serbia) were sampled during autumn 2017. Th e sampling strategy was to sample two strong, one medium and two weak colonies from each apiary. Th e selected colonies were without visible clinical signs of any disease. Using a realtime RT-PCR analysis at least one virus was detected in 87.33% of tested colonies. Th e most prevalent virus was DWV found in 74% colonies, followed by ABPV, SBV and CBPV (identifi ed in 49.30%, 24.00% and 6.70% colonies, respectively). Single virus infection was found in 28.67% colonies (DWV, ABPV, SBV and CBPV in 21.33%, 4.0%, 2.67% and 0.67% colonies, respectively). In more than a half of tested colonies (58.66%), simultaneous infection with more than one virus was detected. In 12.67% of tested colonies, not any of examined viruses was found. With regard to the situation across the regions, DWV had the highest prevalence in all regions (66.70 -83.30%), while the least prevalent virus was CBPV (0 -19%). Except for DWV, the prevalence of the remaining three viruses diff ered signifi cantly between the regions. Th e ABPV was highly prevalent in Vojvodina Province and in Eastern Serbia, but with a low prevalence in Central Serbia. Also, SBV was highly prevalent in Vojvodina Province and mainly low-prevalent in other regions. Th e highest prevalence of CBPV was detected in Southern Serbia, than in Eastern Serbia, quite low prevalence (about 5%) was detected in Central and Western Serbia, and the virus was not detected in apiaries from Vojvodina Province. No signifi cant diff erences were found in the prevalence of DWV, ABPV, SBV and CBPV infections between weak, medium and strong colonies (Ćirković et al., 2018). For genetic relationship and phylogenetic analysis partial coding sequences encoding a part of polyprotein gene of DWV, a capsid protein gene of ABPV, a part of RNA-dependent RNA polymerase (RdRp) gene of CBPV, and a part of polyprotein gene of SBV of detected virus strains in Serbia were used along with related sequences deposited in NCBI GenBank. Th e similarities found for 9 DWV strains from Serbia with deposited nucleotide sequences in NCBI GenBank were 99 to 98%. Some of the nucleotide sequences of Serbian DWV strains were highly similar to those from United Kingdom, but most of them were similar to virus strains from diff erent countries, which is in line with the hypothesis of relatively recent evolutionary diversifi cation of DWV and its worldwide distribution. Th e similarity between nucleotide sequence of ABPV strain from Serbia and those deposited in NCBI GenBank was between 93 and 97%. Th e Serbian ABPV strain was distant from those from north and western European countries, and the highest similarity was found with Hungarian ABPV strains. Th is may be explained by the geographical vicinity and trade between beekeepers of the two countries. Th e Serbian SBV strains were mostly similar to those from diff erent European countries and Russia, and more distinct from strains from other parts of the world. Th e similarity between nucleotide sequence of Serbian CBPV strain and other deposited in NCBI Gen-Bank was only from 93 to 96% indicating probable distinct virus evolution in diff erent geographic areas and the need for more studies that should be done to obtain clearer picture (Ćirković et al., 2018) Milićević et al. (2018) tested 30 honey samples from Serbian apiaries originating from the local markets sampled from 12 diff erent districts in central and northern Serbia, as well as 40 samples of both honeys and live and dead bees, from four apiaries without any visible health disorder, located in central Serbia. Th e samples were tested for the presence of BQCV, KBV, DWV, ABPV, SBV, and CBPV by multiplex RT-PCR. Out of all six examined viruses, only BQCV genome was found in 24 honey samples (80%) from the local markets. Also, BQCV was the only virus found both in bees and honey samples from tested apiaries. Th e virus was detected in three out of four apiaries, and the virus prevalence at a hive level was between 86.6 and 100%. Th e partial polyprotein coding region of detected BQCV strains were sequenced. Th e phylogenetic analysis showed that Serbian BQCV strains were similar to each other (98.5%), and mostly similar to some other European strains (86.4%) clustered by geographical origin. Th e high similarity observed between Serbian and Hungarian isolates that clustered together was found to be quite expected, due to the close trading and communication between two neighbouring countries, the same or similar habitat, and pasture. Th e most divergent virus strains were those from the USA (Milićević et al., 2018).
Th e testing of bees samples taken from diseased or dead honey bee colonies provided by beekeepers all over the Republic of Serbia to the Department of Biology, Faculty of Veterinary Medicine, on the presence of viruses in the fi ve-year period (2014 -2018) showed that the prevalence was between 73.12 and 87.16% for DWV, 61.54 and 81.45% for ABPV and between 58.82 and 64.22% for CBPV, depending of the year of testing (Stanimirović et al, 2019).
In addition to previously described studies, Tarić et al. (2019) investigated the diff erences between the presence of virus infections of the brood and adult bees depending of the used apiculture technology (conventional vs. traditional). Th ere are certain regions in the Republic of Serbia where bees are still kept in a traditional way, in primitive hives made of wicker -so-called trmka hives. Th e study was conducted on the Pešter Plateau (1,059 km2), Raška District, southwest Serbia. Samples of bees from 144 asymptomatic honey bee colonies (120 kept in commercial hives and 24 colonies kept in traditional trmka hives) originating from 18 apiaries (15 commercial apiaries and 3 apiaries with trmka hives) were tested for the presence of the ABPV, CBPV, DWV and SBV by molecular diagnostic methods. Th e SBV was detected in 96.67% samples from commercial colonies and in only 33.33% samples from traditional hives. In addition, the occurrence of viruses in adult bees was signifi cantly higher in commercial colonies. Th e ABPV was found in 96.67%, and CBPV and DWV were detected in 100% of bee colonies for commercial purpose, contrary to the 33.33% colonies detected positive for ABPV, CBPV and DWV in traditional apiaries. Only one out of three apiaries with traditional hives was found positive for one or more viruses. Th e obtained data showed that in the brood and adult bees without clinical symptoms reared in a traditional way in primitive hives, the prevalence of all monitored viruses was up to 33.33%, which is within the limits of normal distribution of viruses in bee colonies in natural conditions. Moreover, traditional trmka hives provide signifi cantly better conditions for maintenance of bee health and their resistance to pathogens (Tarić et al., 2019).
Th e results of all mentioned published studies are summarized in Table 1.  (4), in which some of bee viruses (CBPV and SBV or ABPV, CBPV, SBV and DWV) were not detected, in all other studies, all tested viruses (ABPV, CBPV, SBV, DWV and BQCV, except CWV) were detected at a high percentage. In these studies, ABPV was present in a range from 7.14 up to 90% of samples, CBPV from 3.3 up to 100%, SBV from 3.3 up to 96.67%, DWV from 3.3 up to 100% and BQCV from 50 up to 100% of tested honeybee samples. CWV presence was tested in only one study (Kulinčević et al., 1990) in small number of samples and found to be present in up to 20% of the honeybee samples. Only KBV was not detected in any of tested samples and studies in Serbia (Table 1). In studies outside of Europe, that is, in diff erent regions of China, in a study conducted in the period 2011 -2012, BQCV was reported the prevalent virus (90 -100%), followed by DWV detected in 40 -100%, SBV in 0 -100% and IAPV found in 0 -60% of honeybee samples. CBPV and ABPV were detected only sporadically, and KBV was not detected at all (Ding et al., 2016). In diff erent regions of Argentina, in a study conducted in the same period, DWV was found to be most prevalent ranging from 9 to 90%, followed by BQCV found sporadically in up to 60% of the honeybee samples. ABPV and CBPV were found only sporadically 0 -40%, and SBV, KBV and IAPV were not detected at all (Ding et al., 2016).

CONCLUSION
Th e presence of bee viruses has been studied in the Republic of Serbia for more than 35 years, yet there are only a few studies performed and published mainly in the past 5 to 10 years. High prevalence of diff erent bee viruses identifi ed in Serbian apiaries in most of the studies may be partly explained by high density of bee colonies in geographic regions from which bee samples were collected. Also, the intensive and uncontrolled trade and transport of bee colonies, queens, or equipment could be one of the reasons for the transmission and spread of these viruses between apiaries. Th e tradition of beekeeping and intensive beekeeping practice in Serbia, as well as many diff erent challenges infl uencing the health of bee colonies impose the need for more comprehensive and continuous studies on virus presence and infl uence on bee health. Viruses, already present in apiaries in the Republic of Serbia, in diff erent aspects of decreased immunity in bees, could become the ultimate extinction factors for bee colonies. Th e knowledge on virus's presence and prevalence in apiaries will infl uence the existing trading practices and better protection of healthy apiaries, and improve the management of infected apiaries in the Republic of Serbia.