In the current context of worldwide honey bee colony losses, among which the varroa mite plays a major role, the hope to improve honey bee health lies in part in the breeding of varroa resistant colonies. To do so, methods used to evaluate varroa resistance need better understanding. Repeatability and correlations between traits such as mite non-reproduction (MNR), varroa sensitive hygiene (VSH), and hygienic behavior are poorly known, due to practical limitations and to their underlying complexity. We investigate (i) the variability, (ii) the repeatability of the MNR score, and (iii) its correlation with other resistance traits. To reduce the inherent variability of MNR scores, we propose to apply an empirical Bayes correction. In the short-term (ten days), MNR had a modest repeatability of 0.4, whereas in the long-term (a month), it had a low repeatability of 0.2, similar to other resistance traits. Within our dataset, there was no correlation between MNR and VSH. Although MNR is amongst the most popular varroa resistance estimates in field studies, its underlying complex mechanism is not fully understood. Its lack of correlation with better described resistance traits and low repeatability suggest that MNR needs to be interpreted cautiously, especially when used for selection.

The western honeybee Apis mellifera exhibits a diverse set of adaptations in response to infestations by its most virulent disease-causing agent, the ectoparasitic mite Varroa destructor. In this study, we investigated the effect of honeybee pupae genotype on the expression of four host and parasite traits that are associated with the reproductive phase of the mite in the brood of its host. We first phenotyped cells containing bee pupae to assess their infestation status, their infestation level, the reproductive status of the mites, and the recapping of cells by adult workers. We then genotyped individual pupae with five microsatellites markers to compare these phenotypes across full sister groups. We found that the four phenotypes varied significantly in time but did not across the subfamilies within the colonies. These findings show that V. destructor mites do not differentially infest or reproduce on some particular honeybee patrilines, and that workers do not target preferentially specific pupae genotypes when performing recapping. These findings bring new insights that can help designing sustainable mite control strategies through breeding and provide new insights into the interactions between A. mellifera and V. destructor.