The detection of Israeli Acute Paralysis virus ( IAPV ) , fipronil and imidacloprid in professional apiaries are not related with massive honey bee colony loss in Spain

This study is presented in order to determine the presence and impact of some factors related with honeybee colony losses such as Israeli Acute Paralysis virus (IAPV) and certain pesticides. Samples (house worker bees and stored pollen from brood chamber) were selected from one hundred apiaries, half of them with a clear underpopulation, in accordance with region and time of the year. Total prevalence of IAPV either in spring or in autumn was 18% (CI95% = 9.9-26.0; p < 0.0001), no relationship between IAPV and depopulated colonies was established. Fipronil was only detected in two samples of stored pollen from asymptomatic colonies and imidacloprid was not detected in any sample. Like IAPV, neither fipronil nor imidacloprid appeared to be directly related with the generalized problem of colony death and honey bee losses in professional Spanish apiaries. Additional key words: Apis mellifera, honeybee colony loss, neonicotinoids, pesticides, viruses.

problem has become more widespread across the entire Spanish territory.Some hypotheses regarding the origin of this phenomenon have been linked to the deleterious effect of certain pesticides (Chauzat et al., 2006), the possible introduction of a novel infectious agent, such as Nosema ceranae (Martín-Hernández et al., 2007;Higes et al., 2008;revised by Williams, 2009) or of the Israeli Acute Paralysis Virus (IAPV: Cox-Foster et al., 2007), as well as other aetiologies (e.g., environmental and climatic changes, radiation, the synergistic action of different agents).However, these latest hypotheses have not always been supported by well founded scientific data.
In this study, a total of 100 samples were included in the study from 100 different professional apiaries from 33 different Spanish provinces.The case definition for the disease was: the unusual presence of few adult bees from the hives with no or little builds up of dead bees in or in front of the colony compared with highly productive colonies in the same region and time of the year.
Each sample was composed of adult house worker honey bees (n > 200) and stored pollen (> 100 g) from brood chamber combs.The samples were collected by the veterinary services of beekeeping and included a report about the sanitary status of the colonies in the moment of sampling.Samples were obtained in the spring (n = 50) and autumn (n = 50) of 2006 (Figs.1a  and 1b) and while 55 of them came from diseased colonies (n = 28 spring samples, n = 27 autumn samples), the other 45 were from asymptomatic colonies (n = 22 spring samples, n = 23 autumn samples).
IAPV analyses were carried out as described previously (Maori et al., 2007) in the virus laboratory of Hebrew University of Jerusalem (Israel).The presence of f ipronil and imidacloprid in stored pollen were carried out as described previously (Jiménez et al., 2007;Higes et al., 2008).
To define the correlation between the presence of each one of these three agents and colony depopulation, three successive and independent contrasting hypotheses were proposed based on the Chi-squared asymptotic test and calculating the exact probabilities.The statistical analyses were all performed using SPSS 15.0.1.
Total prevalence of IAPV in the period of sampling was 18% (CI95% = 9,9-26,0; p < 0.0001), and 9 apiaries from 8 different Spanish provinces were IAPV positive in spring.Similar results were obtained in autumn 2006, when 8 apiaries were found as IAPV positive (Figs.1a and 1b).No significant correlation was established between IAPV detection and depopulation symptoms (Chi-square, p > 0.05, Table 1).Fipronil was only detected in 2 samples of stored pollen from asymptomatic colonies in two provinces during spring (Fig. 1a), and it was not related with colony depopulation (Chi-square, p > 0.05).Imidacloprid was not detected in any of the stored pollen samples.
IAPV was first described in Israel (Maori et al., 2007) and it was associated with bee mortality, even though only some hives in an apiary were affected.It was suggested that genetic diversity among honey bees could explain the different susceptibility/resistance Honeybee colony loss in Spain 659 between hives in an apiary.Thus, it was proposed that IAPV is related to the problem of bee colony losses in the USA (Cox-Foster et al., 2007) and that it is a significant marker for Colony Collapse Disorder (CCD).However, further works found IAPV to be present in U.S. bees since prior to CCD, and that U.S. isolates of IAPV are not substantially different from those found in Israel (Chen and Evans, 2007).Indeed, no association was established between this virus and the severe winter losses in France, since the prevalence of IAPV in this country is 14% (Blanchard et al., 2008), indicating that other pathogens/factors may be responsible for the colony loss.In our research work we found no relationship between IAPV and depopulated colonies, our results are consistent with these latter conclusions.Indeed, the prevalence of IAPV (18%) is very low compared with other pathogens that are more prevalent in Spain such as Nosema ceranae (Higes et al., 2006;Martín-Hernández et al., 2007) and Varroa destructor (Bee Pathology laboratory, Centro Apícola Regional, unpublished data) do appear to affect the health status of our apiaries.Thus, IAPV does not seem to be a good marker for depopulation and colony loss, although the exact pathogenic role of this virus in Spanish honey bee colonies remains to be defined.It has been proposed that bees come into contact with pesticides through the pollen and/or nectar they consume.Accordingly, stored pollen could be a chronic source of toxic to the bees (Faucon et al., 2005), which would mainly exert their influence in autumn/winter since stored pollen is practically the only pollen that the bees consume in this season (Seeley, 1985).The toxicity of sub-lethal doses of fipronil in honey bees has been demonstrated in laboratory studies but until now, not in field conditions (El Hassani et al., 2005).In Spain, its use has been authorized for sunflower (Helianthus annuus) and other crops since 2004.As a consequence, its application to sunflower seeds is very recent and it is limited to very specific geographic sites in central and southern Spain.Indeed, in 2007 fipronil residues were only detected in two samples but they were not directly related with unhealthy colonies.Imidacloprid is a neonicotinoid that is toxic to honey bees, and authorized in Spain to treat corn (Zea mays) crops, fruit trees and vegetable crops but their prevalence in stored pollen samples from Spain is very low as described previously and not related with colony losses from professional apiaries (Higes et al., 2009b).Actually the bee losses observed by beekeepers in France or Belgium are not attributed to imidacloprid and other causes such as pathogens or other unidentified synergistic factors have been considered in the origin (Schmuck, 2004;Faucon et al., 2005;Nguyen et al., 2009).
Thus, like IAPV, neither of these two insecticides we tested appears to be directly related with the generalized problem of colony death and honey bee losses in professional Spanish apiaries.).Honey bee losses are likely to be provoked by different causes but the recent pandemic produces weakness, depopulation or death clinical features similar to those caused by the emergent pathogen N. ceranae, known to be a shortterm bee pathogen (Higes et al., 2007;Paxton et al., 2007;Chen et al., 2009a,b) and a long-term colony pathogen (Higes et al., 2008(Higes et al., , 2009a;;Korpela, 2009).Along similar lines, well-known pathogens like V. destructor, one of the main pathogens of A. mellifera, can cause a similar illness and act as a re-emerging agent if not suitably controlled.In our country, the combination of both pathogens and the highly pathogenic viruses associated with them could increase the risk of death in infected colonies, without pesticides exerting a significant effect (Higes et al., 2009b).

Figure 1 .
Figure 1.Collection sites of the Spanish apiaries monitores in a) spring 2006 (N = 50) and b) autumn 2006 (N = 50).• Samples negative to both Fipronil and IAPV.L Samples positive to Fipronil.F Samples positive to IAPV.None sample was positive to both of studied factors.

Table 1 .
Frequencies of detection of studied factors (IAPV and Fipronil) in colonies showing depopulation symptoms or asymptomatic