Skip to main content

Honeybee communication: animal social learning at the height of social complexity

Periodic Reporting for period 4 - BeeDanceGap (Honeybee communication: animal social learning at the height of social complexity)

Okres sprawozdawczy: 2020-08-01 do 2021-01-31

The honeybee (Apis mellifera) waggle dance is one of the most complex and celebrated communication systems in the animal world. Foraging bees use an abstract symbolic code to communicate the distance, direction, and quality of resources to their nestmates, and the information network that results has been shaped by natural selection to produce fine-tuned, rapid, group-level responses from colonies that can contain up to 80 000 individuals. BeeDanceGap addresses two fundamental gaps in our current understanding of this extraordinary behaviour: (1) how is it that bee brains process the complex information provided by the dance (2) why have they evolved to do so?

To address the first of these questions, we assayed the changes in gene expression that occur within the bee mushroom bodies- an area of the brain that is associated with learning and memory- immediately after following a dance. We identified genes that respond to the acquisition of spatial information through the dance, and compared them to those that are involved in acquiring the same information through individual experience. We found that Apidaecin was consistently expressed in the same way through both mechanisms.Therefore there is suggestive evidence that points at a role for apidaecin in the regulation of cognitive functions in the honeybee mushroom bodies associated with foraging behaviour.

To address the second question, we assayed the importance of dance-following for food discovery under varying forage distributions. Foragers have multiple information sources available to them in the nest, and we developed and used network-based diffusion analysis models to evaluate the relative contribution of each pathway to real-world food discovery. We found that the waggle dance is critical to the discovery of new food sources when colonies shift focus from one patch to another of the same type, but may be far less critical when bees revisit food sources or shift between species. A better understanding of the extraordinary social communication system of these remarkable insects extends our knowledge of sophisticated collective behaviour, but it also provides insight into a critical information pathway that drives one of the world’s most important and threatened pollinators to food. Our research is important not simply because honeybee foraging behaviour is an evolutionary marvel, but because it is a significant contributor to pollination services at a global scale.
The primary axis of the project focused on the consequences of spatial learning through the dance signal, for neural gene expression. The first step was been to identify genes that are differentially expressed when bees learn through their own experience about food locations at different a) distances b) directions. We developed a protocol through we could manipulate an individual’s perception of the distance and direction that she has flown, and through this manipulation, identified a set of genes that are differentially expressed in the bee mushroom bodies when bees perceive themselves to have flown long or short distances. This provided a candidate set of genes that form the basis for the all-important second step in the project, which was to search for similar gene expression differences in bees that follow dances for such locations. We allowed bees to forage for feeders at long or short-distances and perform dances, sampling these individuals and (critically) those bees that followed the same dances. This identified a particular gene- Apidaecin- that responded to distance in bees that flew long/short distances, bees that perceived themselves to have done the same, and bees that learnt about those distant/close locations through dances. We have disseminated these results through several conference symposia, and are in the process of publication.

The second axis of the project focused upon the use of network-based diffusion analysis (NBDA) to compare the contribution of the different information pathways that are available to honeybees. The core assumption underlying NBDA is that if social transmission is occurring, then the spread of a novel behaviour should follow a social network that reflects opportunities for information transfer between individuals. However, earlier iterations of NBDA had several limitations. The approach that we developed, in order to achieve our primary aim of assaying information flow through bee colonies, built on this by incorporating multiple dynamic networks for comparison within a single model. We published this tool as an R-package and disseminated further through an accompanying extensive “How To” paper to encourage uptake by other ecologists, and through workshops at international conferences. We went on to apply our approach to our own empirical data, comparing the importance of three different honeybee communication systems as drivers of forage discovery. We showed that the waggle dance is critical to the discovery of new food sources when colonies shift focus from one patch to another of the same type, that it is not important for food rediscovery, that reliance on it does not vary with distance forage and that it may be less important for shifts between forage types. These results indicate that within-species forage shifts may have been key to the evolution of this unique communication system, and each has been presented at multiple conference symposia and is either published or in the process of publication.
The incorporation of multiple dynamic networks into established NBDA techniques is a development with potential applications for the study of rapid information flow in any social group, and we expect that our tools will be applied to expand the use of NBDA to study social learning and communication within ecology. Within our own system, our results have allowed pinpointing of the circumstances that may have been critical in dance evolution, and thus make headway towards understanding the ecological selection pressures that drove the evolution of this unique communication system that is full of redundancy.

We have introduced next-generation transcriptomic approaches to the study of animal social learning- a psychology-based subject area that rarely overlaps with molecular biology. Through this, we have shown that even very brief signal exposure (following dance circuits over the course of a single morning) has a molecular signal at the transcriptomic level.