Honey bee invaders show us balancing selection in action
Australia failed to stop an exotic honey bee invasion but, in trying, succeeded in collecting a dataset that explains why such invasions are so resilient.
As anyone who has ever travelled to Australia knows, we take our biosecurity very seriously. Invasive species have extracted heavy tolls on the continent’s ecosystems and agriculture. So when a swarm of slim, darkly-banded honey bees was spotted hanging from a ship’s mast in the tropical port of Cairns in 2007, the danger was immediately understood. This was Apis cerana, a native of Asia with a history of invasiveness and the potential to transfer pests and pathogens to resident pollinators. An eradication program was launched, which saw the public get involved with gusto by reporting nests and swarms. Over the years, hundreds of colonies were sampled and destroyed. Nevertheless, despite these efforts, A. cerana multiplied and spread: the battle was lost. Today we estimate the population at around 10 000 colonies, possibly many more.
What the eradication program left behind though was a beautifully complete genetic timeline of a successful social insect invasion. Hymenopteran social insects (all ants, some bees and wasps) number among the world’s nastiest invasive species, yet their success as invaders is counter-intuitive because they should be vulnerable to the genetic bottlenecks that typically accompany invasions. This is because the Hymenopteran sex determination mechanism relies on heterozygosity at a single locus to produce females, and high genetic load at this sex locus should choke off small founding populations. Until now, a major impediment to understanding how invaders overcome this problem has been the lack of data from natural invasions. A. cerana’s invasion of Australia provided a rare exception to this generality and a chance to shed light on the puzzle. Was genetic load at the sex locus following a founder event as severe as predicted? And did it persist over time?
Image: Apis cerana swarm in the invasive Australian population (photo: R. Gloag).
To supplement Biosecurity’s historical samples we made our own collections in 2014-15, again with the help of local authorities, beekeepers and the residents of tropical Queensland who reported nests to us and helped us extract them from every imaginable nook and cranny – farm equipment, letterboxes, stereo speakers, birdboxes, the tyre of a forklift, even the belly of a garden Buddha. There are few better ways to spend a winter than traveling tropical Australia collecting bees. These samples were combined with over a year of hard grind and trouble-shooting in the lab until we eventually had our answer. Austraila's invasive A. cerana population has seven sex locus alleles (c. one-third of native-range populations), initially present in highly unequal frequencies. Rapid balancing selection, in the form of negative frequency dependent selection, acted to overcome genetic load at this crucial sex locus by driving allele frequencies towards equilibrium and restoring maximum heterozygosity. Thus while genetic load was very high for early colonies, this cost was short-lived. Stochastic models then allowed us to illustrate that our empirical data were in line with the predictions for balancing selection, and that the bee’s social life history drives selection at the sex locus in two ways: not only are homozygotes lethal, but they also reduce the female workerforce and thus a colony’s ability to rear brood and reproduce.
The end result is, we feel, a compelling case study of how a social insect biological invasion can unfold, and provides the first empirical evidence that balancing selection can have dramatic implications for the ecology of a species. If you ask us, its one for the textbooks.
There of course remain open questions. The genetic footprint of Australia’s A. cerana population is consistent with just one single founding queen. If that is the case, what needed to happen in those very first few generations to ensure adequate genetic diversity at the sex locus made it into reproductive castes and thus available for selection to act upon? The answer lies in better understanding A. cerana’s reproductive behaviour.
There has also been a new development since our study was completed: a second invasion of Apis cerana in Australia, this time in a different tropical port 400km south of the existing population. We are working closely with Queensland Biosecurity to assist in the new eradication program and so far it's going well. Only time will tell whether history is repeating itself and we see balancing selection once again act to keep essential diversity at the sex locus high.
Image: Apis cerana nesting inside a birdbox in suburban Cairns, Far North Queensland Australia (photo: R. Gloag)
Image: Ben Oldroyd and Ros Gloag on the hunt for invasive honey bees in the cane fields of tropical Queensland (photo: W. Scougall).
Image: An Apis cerana swarm in its invasive range in tropical Australia.
Our article in Nature Ecology & Evolution can be found here: http://go.nature.com/2fhcAb3