A blog post highlighting the article by R. Gloag, K. Tan, Y. Wang, W. Song, W. Luo, G. Buchman, M. Beekman, B. P. Oldroyd in Insectes Sociaux

Written by Ros Gloag

Some social insects have proved to be adept invaders. Assisted by the international trade of the modern world, these species have spread far beyond the ocean and mountain barriers that once determined their distributions. In some cases, these range expansions have brought previously isolated sister species back into contact. What happens when such species try to mate?

We were interested in this question of interspecific mating in the case of two honey bees: the Western honey bee Apis mellifera and the Eastern honey (or hive) bee, Apis cerana. These species diverged from a common ancestor at least 6 million years ago, with A. mellifera native to Europe and Africa and A. cerana native to Asia and India. Western honey bees have of course since been transported, in association with agriculture, to every human-inhabited continent on earth. Eastern honey bees meanwhile, have been quietly expanding their range too in recent decades, invading both Papua New Guinea and Australia. Thus what were allopatric (or separate) ranges for millions of years have suddenly become partially sympatric.

A swarm of Apis cerana hangs from a branch in its invasive range of Northern Australia, where the species has recently come into contact with A. mellifera. The newly-mated queen will be concealed at the centre of the swarm: but who did she mate with?

The possible outcomes of A. mellifera and A. cerana mating are varied. It may produce high-fitness hybrids, low-fitness hybrids or no viable offspring at all. In the case of honey bees, there is also a more unusual possibility; interspecific mating might cause queens to produce some female diploid offspring asexually via a process called thelytokous parthenogenesis. Thelytoky is not uncommon in Hymenoptera, though the mechanisms controlling it vary between species. In honey bees, it appears to have some genetic basis, but its unclear whether environmental factors – such as interspecific mating – also play a role in determining its incidence. Honey bee queens mate with twenty or more males during a short period early in their lives and store the sperm, so it is unlikely that naturally-mated queens will have mated exclusively with the wrong species. As such, any peculiar effects of interspecific mating could be easily obscured in populations where the two species co-occur.

We decided to perform an experiment to reveal the effects of interspecific mating on the offspring of A. mellifera and A. cerana. We performed reciprocal crosses via artificial insemination (inseminating queens of each species with the sperm of the other species) in China. Artificial insemination is a fairly standard beekeeping procedure for A. mellifera, but a much trickier business for the relatively diminutive A. cerana. Enough inseminated queens survived the procedure though to confirm that theytoky is not a consistent outcome of these matings. We detected only the odd few thelytokous eggs, from both queens and laying workers. Rather, our results confirmed that interspecific mating has fitness costs for both species: cross-inseminated A. mellifera queens produced only males or inviable hybrid females, while cross-inseminated A. cerana queens produced either males only or no eggs at all. Interestingly, A. cerana workers sometimes rebelled against their “wrongly-mated” queen and took control of reproduction themselves by laying unfertilized male-destined eggs.

Of course, understanding what happens if species mate is different to knowing whether they do mate. A previous study confirmed that A. mellifera will sometimes mate naturally with A. cerana males, but whether the reciprocal pairing ever occurs is unknown. We checked the sperm-storage organs of 17 A. cerana queens collected from Australia’s invasive population and failed to detect A. mellifera semen, despite the fact that we have observed A. mellifera males hanging about in areas where Australian A. cerana queens mate. Possibly A. cerana queens simply cannot survive interspecific matings with their larger sister species, which would be a particularly brutal and conclusive form of reproductive interference because its effects could not be diluted by multiple mating.

Wherever interspecific mating does occur between Western and Eastern honey bees, we can expect that natural selection will eventually intervene. After all, there are other honey bee species in the world that naturally coexist without incident, generally by having species-specific mating times and locations. A. mellifera and A. cerana are recent bedfellows, but given that interspecific mating in their case appears to have no redeeming features, selection should act to favour those queens and drones that succeed in keeping sex strictly within the species.