Differential infestation of honey bee, Apis mellifera, worker and queen brood by the parasitic mite Varroa destructor

We examined the distribution of Varroa destructor on worker and queen brood in colonies of A. mellifera. With both worker and queen hosts present, the mite prevalence value for worker hosts was 75.0 $\pm$ 4.0% (lsmean $\pm$ SE), compared to 5.1 $\pm$ 4.0% for queen hosts ($P < 0.0001$). We also examined the response of mites to cuticular extracts of 5th instar worker and queen larvae using arrestment bioassays. In binary-choice tests at 0.5 larval equivalents (Leq), worker extract arrested 84.79 $\pm$ 4.98% of the mites, while queen extract arrested 15.21 $\pm$ 4.98% ($P < 0.0001$). At 0.8 Leq, worker extract arrested 89.75 $\pm$ 4.98%, while queen extract arrested 10.25 $\pm$ 4.98% ($P < 0.0001$). We also measured the repellent activity of royal jelly extract in a repellent bioassay. Royal jelly extract repelled 78.5 $\pm$ 2.6% of mites at 5 mg royal jelly equivalents (Rjeq); 85.6 $\pm$ 3.7% at 10 mg Rjeq; and 89.2 $\pm$ 3.8% at 20 mg Rjeq. The response at each dose was greater than the 10.5 $\pm$ 2.9% mites repelled by solvent controls ($P < 0.0001$). Our findings suggest that the low incidence of mites in queen brood is due, in part, to the repellent activity of royal jelly, and possibly to intrinsic differences between larval chemistries