Abstract
European honey bees (Apis mellifera) pollinate approximately one third of the world’s fruits and crops. The current loss of honey bee colonies is thus a major threat to the agricultural industry, to food production and to global economies worldwide. Varroa (Varroa destructor) and viruses associated with this parasitic mite, in particular deformed wing virus (DWV), have been identified as probable causes of colony losses.
Some honey bee colonies can survive varroa infestations, especially colonies that display hygienic behaviour targeted to varroa-infested cells. Worker bees displaying varroa sensitive hygiene (VSH) open capped cells containing varroa, remove the brood and clean the cells. This behaviour disrupts mite reproduction. The origin and nature of stimuli used by VSH bees to detect varroa in capped brood cells remain largely unclear. It is also unclear why some but not all worker bees perform VSH. This thesis identifies features that characterise brood cells targeted by VSH bees. It also examines the hypothesis that DWV infections might affect antennal function, a possibility suggested by evidence indicating that DWV titres in the antennae of bees performing VSH are lower than in the antennae of bees not seen performing this behaviour.
To identify features that characterise brood cells targeted by bees performing VSH, comparisons are made between the contents of two types of brood cell: cells targeted by VSH bees and cells infested with varroa but not targeted by VSH bees. For each targeted and non-targeted cell included in the analysis, two layers of surrounding cells are uncapped and the cell contents examined. Records are taken of the developmental stage of each pupa, its DWV titres, and the mite populations found in each brood cell. The results show a correlation between VSH behaviour and the presence of high numbers of female varroa and mite offspring in brood cells. Cells surrounding targeted brood cells are found to have higher proportions of varroa-infested cells than cells surrounding non-targeted brood. DWV titres in brood targeted for removal by VSH bees and in brood from varroa-infested cells ignored by bees performing VSH are found to be similar.
Molecular, ultrastructural and behavioural analyses are used to explore the possibility that DWV infection might affect antennal function and hence the performance of VSH. Molecular analyses are used to confirm that DWV is present in antennal tissues and to compare DWV loads in the antennae of infested versus non-infested brood, VSH bees versus non-VSH bees, and newly emerged bees with and without symptoms of DWV infection. Ultrastructural methods are used to determine where in the antennae virus particles are located and whether antennal ultrastructure is affected by DWV infection. Water and sucrose responsiveness in VSH bees and in non-VSH bees from the same hive are tested to provide an indirect indicator of antennal function.
DWV titres are shown to be higher in the antennae of varroa-infested brood than in the antennae of non-infested brood. DWV titres are also found to be higher in newly-emerged bees showing symptoms of DWV infection than in asymptomatic bees, but no differences in DWV titre are identified in the antennae of VSH bees versus non-VSH bees. In newly-emerged adults showing symptoms of DWV infection, individual virus particles and structured arrays are identified both in the antennal haemolymph and in basal regions of the antennal sensory epithelium. No virus particles are identified at the level of the antennal sensilla. VSH bees are found to have a lower sucrose response threshold than non-VSH bees, but no difference is detected between the overall sucrose response scores of these two groups, or their responsiveness to water. These results indicate that whether or not worker bees perform VSH is unlikely to be related to DWV infections in the antennae.