The aim of the post-hoc study was to identify critical survival parameters and risk factors (pathogens, parasites, possible residues of active substances dangerous to bees) for colony losses with the aid of retrospective examinations of sample material (samples of bees and bee bread) from colonies that had died or survived the winter. The basis for this was provided by the bee and bee bread samples taken from the living colonies during the stand-visits during the second stand visit of the surveillance study in autumn 2015, as this was the closest time to the wintering period.
From 915 colonies, a complete set of samples was available (autumn: bee sample for varroa infestation analysis, bee sample for pathogen-screening, bee bread sample for residue analyses; spring: bee sample for pathogen-screening) as well as information on overwintering results. From these 915 colonies, 210 colonies (60 dead during the winter and 150 survivors) were selected for the post-hoc study. All 60 dead colonies were included in the evaluations with a complete set of samples in order to obtain a detailed picture of the state of health of the dead colonies shortly before winteringizing. The investigated group of 150 surviving colonies was selected after the winter with the help of a random generator from a total of 855 surviving colonie with a complete set of samples.
The samples were examined for nosema infestation, positive samples were differentiated molecularlbiologically between Nosema ceranae and N. apis.
For the three virus species ABPV, CBPV and DWV a qualitative as well as a quantitative analysis of the samples from the second visit was performed. DWV-positive samples were differentiated into the two virus types DWV-A and DWV-B.
The residue analysis of the bee bread covered a broad spectrum of about 300 analytes and enabled the determination of a possible exposure of the bee colonies to pesticides, biocides, veterinary drugs or pesticides and ubiquitous contaminates sites from the past.
Nosema ceranae was by far the predominant Nosema species. It was present in all 42 positive samples (20 %) from the second visit and all 42 positive samples (26 %) from the third visit. N. apis was only detectable in one of the samples from the second and third visit. There was no significant difference in the proportion of Nosema-positive samples taken in the previous autumn (17% of dead colonies; 21% of living colonies) between colonies that died during the winter and colonies that were successfully overwintered. There was also no significant difference in the amount of spores per bee between dead and living colonies.
As the virus tests of the bee samples from autumn 2015 on ABPV, CBPV and DWV showed, there were considerable differences in the proportion of infected colonies. ABPV was the most commonly detected virus, followed by DWV and CBPV.
For ABPV, there were no significant differences in the proportion of positive samples between dead and surviving colonies. For DWV, the proportion of positive specimens was significantly higher among the colonies who died during the winter than among the surviving colonies. For CBPV, the proportion of positive samples was higher among surviving than among dead colonies. What this unexpected result is due to cannot be deduced from the data of the investigated sample collective.
Multiple infections by more than one of the three examined pathogens or Nosema ceranae occurred in 90 of the 210 bee samples examined. The proportion of dead colonies was highest in each pathogen in mating with DWV and lowest in mating with CBPV. By far the highest winter losses (47%) were achieved in mating DWV and ABPV. Those peoples who had a double infection with ABPV and DWV died significantly more often than those who were only infected with ABPV.
In the residue tests for chemical substances (pesticides, biocides, veterinary drugs, contaminates from the past, etc.) of 210 bee bread samples from individual colonies, no residues were detectable in 42 samples (20%). A total of 48 analytes were detected in the residue-positive samples. The three most frequently detected active substances in descending order were the fungicide fludioxonil (62 positive samples), the synergist piperonyl butoxide (57 positive samples) and the insecticide and acaricide tau-fluvalinate (43 positive samples).
By far, esfenvalerate (22 samples found positive) and thiacloprid (21 samples found positive) followed. In this order, it is surprising that piperonyl butoxide could be detected in 57 samples. Hence, this compound, which is used in combination with pyrethrum or pyrethroids in some plant protection or biocidal products, can be considered as an indicator for exposure, even when pyrethrum was not and pyrethroids were only detected in a small number of the investigated samples.
Of all four active ingredients, which according to the EU directive fall under the partial ban for their use in certain crops and applications, clothianidin and thiamethoxam were not found in any sample, imidacloprid was found in seven and fipronil in two samples. This amounts to an exposure rate among the 210 investigated colonies of 3% found for imidacloprid and 1% for fipronil, respectively. Of the metabolites of these active ingredients, imidacloprid-hydroxy metabolite was found in one sample and the imidacloprid-olefin metabolite was found in three samples. From the clothianidin metabolites, TZMU was found in 13 samples and TZNG in one sample. The fipronil metabolite fipronil sulfone was not found in any of the samples. Occasionally, the following active ingredients that are not or not any longer approved for plant protection were detected in bee bread: bioallethrin, biphenyl, bromopropylate, chlorfenvinphos, DDD, DDT, dichlofluanide, permethrin, propargite, quinalphos and 4,4-methoxychlor. Potential sources for these contaminations may be found in other fields of application (e.g., wood protection, anti-parasitic use or pest control) or from ubiquitious contaminants from the past.
During the investigation period of the post-hoc study, by using a tested multivariate model, Varroa mite infestation rate as assessed in fall 2015 was found to be the primary driver of winter colony losses 2015/16.
For the overall interpretation of the results of the post-hoc study both with regard to the prevalence of pathogens and parasites as well as to residues of chemical substances (plant protection products, biocides, veterinary drugs, ubiquitious contaminants from the past, etc), it has to be taken into account that the data refer only to the investigation period in fall 2015. In the following winter period 2015/16, the overall lowest winter losses since the beginning of data gathering in 2007/08 were registered. To quantify differences between the different years, long-term survey programs would be required. These would be also necessary to assess changes in the balance of differently virulent virus strains over the years and their interdependence with Varroa status and respective Varroa control treatments.
In summary the post-hoc study showed on the whole a positive picture of the health state of the investigated bee colonies in Austria. The bee colonies investigated during the post-hoc study were free of symptoms of intoxication and the majority was also free of disease symptoms.
In the investigated 210 bee samples that were taken in fall 2015, the harmful deformed wing virus (DWV) was significantly less detectable than in earlier surveys and also as compared to other European comparative data. Because a DWV infection correlates strongly with the occurrence of varroosis, these results indicate that the majority of beekeepers of the surveillance study was able to keep Varroa well under control during the bee season of 2015/16. A further sign for this were the low winter losses during this period.
It was not possible to detect a significant correlation between residues of the analysed chemical substances (plant protection products, biocides, veterinary drugs, ubiquitious contaminants from the past, etc) in bee bread in fall 2015 and subsequent winter losses. Pesticide concentrations were found in most cases far below the values found in cases of suspected bee poisoning.
The finding that successful Varroa control constitutes a most crucial factor for overwintering success of bee colonies paves the way to implement improvements in Varroa control through aimed training measures and hence to reduce winter colony losses.