In bee breeding, the long-term use of honeybee queens that are mated to a single male should be avoided, and breeders should stick to classical multi-male insemination of queens. This result was obtained using an individual-centered simulation (simulating individual queens, males, and worker groups), accounting for genetic and phenotypic honeybee specificities, and in which all queens were either mated to only one male each, or to several males each.

Selective breeding using single-drone insemination and selection for parasite resistance

“Western honeybee” colonies (from the well-known Apis mellifera species) suffer from increasingly high mortality, partly due to an ectoparasitic mite called Varroa destructor [1]. This parasite had coevolved with a similar but different species called the “Asian honeybee” (Apis cerana) that the parasite infests but without big consequences. However, when it switched host, starting to infest Western honeybee colonies at the end of the last century, it multiplied abundantly on its new and naïve host, killing colonies rapidly [2]. Since then, resourceful beekeepers have been trying to adapt their breeding stock to this new threat using genetic selection, trying to accelerate honeybees’ and varroa’s coevolution to bring them to a new equilibrium.

How honeybee breeding is different from other livestock

Several specificities distinguish honeybees from other livestock species, complicating the breeder’s task. Males are haploid (meaning they only possess one copy of each chromosome), while females (queens and workers) are diploid. Queens only mate shortly after they are born, storing the sperm of their multiple mates for the rest of their lives. A queen mates with several males, and as a consequence, workers in the worker group always descend from the same queen but from either: unrelated drones, related but different drones, or from the same drone. Because males are haploid, they produce clonal copies of sperm. Workers that descend from the same drone therefore inherit genetically the same spermatozoa and its alleles. Such workers are maternal sisters but paternal twins and are called super-sisters. Lastly, honeybees live in big colonies comprising up to a few tens of thousands of individuals. Traits that interest beekeepers, such as honey yield or varroa resistance, are therefore measured not on individuals, but on colonies. A colony comprises mainly a single queen and its worker group. The colony phenotype can be modelled as resulting genetically from effects expressed by the queen (called queen or maternal effects) and effects expressed by the workers (called worker or direct) effects. However, certain beekeepers have been inseminating queens with the sperm of a single drone, obtaining worker groups made only of super-sisters. This enables them to simplify the genetic makeup of a colony. In turn, phenotypes of these more homogeneous colonies are more contrasted from one colony to the other [3], and the genetic merit of the single haploid male is better estimated. This is particularly relevant when obtaining the phenotypes is difficult and expensive, as is the case for most traits related to varroa resistance.

Research question

It was so far unclear, however, if the mating strategy of inseminating queens with only one male would indeed bring more genetic gain on traits of interest than inseminating queens with several males (as would be usual).

Methods: computer simulation

To investigate the theoretical consequences of breeding with single-drone inseminated queens, we simulated honeybee populations under selection comprising either classical multi-drone (polyandrous) inseminated queens or alternatively only single-drone (monoandrous) inseminated queens (figure 1). To simulate genetic values and colony phenotypes, we developed an infinitesimal model (supposing the phenotypes to result from a large number of genes with small effects) that accounted for honeybee specificities. We simulated different genetic variabilities of the two genetic effects (worker effects and queen effects), as well as different genetic correlations between them, to cover a wide range of possible trait parameters. We also varied the breeding strategy, choosing queens across all possible candidates (mass selection), or restricting the choice to one queen per maternal sister-group (within family selection).

Results and interpretation

Results showed that both strategies brought strong improvements in genetic effects expressed by workers. Monoandry brought much more gain on worker genetic effects than on queen genetic effects. This can be interpreted by the fact that monoandry permitted to better discriminate, between colonies, the superior drones in terms of worker effects. Indeed, their genetic worker effects were more clearly identified when expressed in the more homogeneous worker groups (where all workers inherited the same drone's alleles). However, because this strategy brought much more modest gains on queen genetic effects, overall gains on colony performance were always superior in the polyandrous mating scenarios. Furthermore, the total number of reproductive males was much lower with monoandry, resulting in a steeper decrease in genetic variability (figure 2), which in turn limited genetic gain after some generations.

Conclusions

Overall, using monoandry as the sole mating strategy for a whole breeding population for 20 years cannot, for purely genetic reasons, be recommended in general over a strategy using classical polyandrous queens. However, this strategy might still be useful if applied only to a limited number of generations, on a subsample of the breeding population, or to phenotype relatives of candidates to selection, for example.

References

1. Traynor KS, Mondet F, Miranda JR de, Techer M, Kowallik V, Oddie MAY, et al. Varroa destructor: A Complex Parasite, Crippling Honey Bees Worldwide. Trends in Parasitology. 2020 Jul 1;36(7):592–606.

2. COLOSS: Prevention of honey bee COlony LOSSes [Internet]. COLOSS. 2024 [cited 2024 Apr 3]. Available from: https://coloss.org/

3. Harbo JR. The value of single-drone inseminations in selective breeding of honey bees. Apiculture for the 21st Century. Cheshire, CT. Wicwas Press. 1999;1–5.