Beyond bee collective decision making: how foraging distance distributions reveal that honeybee waggle dance recruitment varies with landscape

Honeybees communicate the location of flowers through the waggle dance. But how much a colony uses the waggle dance in their decision making and foraging is not known. In a paper in Communications Biology we present a method to quantify how much a colony uses the waggle dance in their decision making and foraging. We found that waggle dance use, and therefore the use of collective decisions, varies considerably and is used more in complex landscapes. This is an edited version of a "behind the paper" blog that appeared on the Springer Nature Research Communities website.

The honeybee waggle dance is a celebrated example of animal communication. It is well-known how bees convey information about the location of food sources through the waggle dance. Perhaps less well known is that the waggle dance also functions as the basic component of a computational device through which the colony decides which resources to focus on. The part of a hive on which the dances are performed is known as the dance floor, and it is through the dance floor that a colony can assess which resources in its environment are best to visit. The dance floor acts as the brain of the hive.

The way this works is both simple and amazingly efficient: bees that have foraged and found a resource, like a patch of flowers, dance for this resource on the dance floor. The better the resource the more often they repeat the dance (honeybees measure what is best in terms of the net profitability of the patch, which discounts travel time). Nestmates that follow dances on the dancefloor are therefore disproportionally directed towards resources with a high profitability. If these nestmates visit such a resource, after their visit, these recruits will also report on the dance floor and repeat the dance in accordance with how they assessed profitability. Through this mechanism the information is filtered and amplified and the most profitable resource voted to the top with an algorithm —somewhat similar to the algorithm that ranks Reddit posts. Through this algorithm a collective decision is made without the need for a central decision maker.

These bees have foraged successfully, as can be seen from the pollen that they carry on their legs. After landing the bees will enter the hive and report the location that they have visited on the dance floor. (Photo by Matthew Hasenjager.)

It has been a matter of some debate to what degree honeybee colonies benefit from collective decision making. This debate has been sparked by experiments in which hives are placed on their side, so that the bees cannot use the angle of the dance with the vertical, which the bees use to give a bearing to dance followers to navigate to the resource. Putting the hive on its side renders the dances therefore meaningless. In most cases such experiments made foraging less effective, but in some cases foraging efficiency increased, indicating that dancing must be energetically costly. But if, and under what conditions, dancing is used by honeybee colonies remains unclear.

We know that most bees perform the dance, provided the resource they report is profitable enough. But what we don't know is how much a colony uses the information that is generated on the dance floor; we don't know how often bees take their directions from dancers. The filtered information on the dance floor will tell the foragers what currently is the most profitable resource that the colony has found, but this does not guarantee that foragers will use this information: they might well ignore go out and simply search for resources themselves.

In our paper we describe a way to measure how much a colony uses the information on the dance floor. To do this, we first observed in simulated honeybee foraging that the distance distributions of resources of bees that look for food without first receiving directions from dances is different from collectively foraging bees. The distance distributions of these foragers that do it on their own are determined by the chance of encountering a first resource on linear search path. Collective foragers, who are directed to the best resources through dances, can explore the area much better and tend to locate the best resource located in a circle around the hive. Based on these different distributions we devised a statistical tool to determine what fraction of bees forage collectively. This gives a measure of how much the dance information is actually used by the colony. We could then compare this to the actual distances visited by foragers in real hives. We do this by filming dances on the dance floor, which allows us to eavesdrop on the communication and work out at what distance the resources are that the bees tell each other about, and how these distances are distributed.

The results were revealing. First, we observed that the fraction of bees that forages collectively is not fixed but highly variable: for some colonies foraging was almost entirely collective, for others it was largely individual and for many it was something in between. This indicates that the amount of foraging that is informed by dances differs between hives, probably because of differences in the surrounding landscape. Secondly, we determined the landscape features that correlated with the use of dance information. We found that in landscapes which were dominated by large, homogeneous areas, such as arable crops, collective foraging was not used much. In landscapes with more heterogeneity, for instance those dominated by gardens, embankments and unmanaged land, collective foraging was used frequently. This result makes sense: these are the environments in which bees will find patches of flowers through various seasons. And we could infer this just from the dances that we filmed on the dance floor in a hive.

Our results shed a light on how bees use the information that they generate on the dance floor and the conditions under which this information is used. Even though under some circumstances the information from the dance floor is little used, in other circumstances it is used extensively. Importantly, the use of dance information is flexible. Therefore, to argue that in man-made environments the dance is maladaptive because it is sometimes not used much would be too simplistic.

It is assumed that the waggle dance evolved in the tropical rainforests of Asia, but it is still an open question which ecological conditions favoured the evolution of waggle dance behaviour. By having a method which allows us to establish to what degree the waggle dance information is used —and therefore has the potential to be useful— we created the means to identify the environmental features, such as high floral diversity or abundance, that may have been important at its origin. And on a more contemporary note, it will allow us to assess and better predict the effects of anthropogenic changes in our environment on honeybee foraging and pollination.

Vincent Jansen and Elli Leadbeater, October 2024

Further reading:

The reference to the paper is

• Palmer, J., Samuelson, A.E., Gill, R.J., Leadbeater, E. & Jansen, V.A.A. (2024) Foraging distance distributions reveal how honeybee waggle dance recruitment varies with landscape. Commun Biol 7, 1306. https://doi.org/10.1038/s42003-024-06987-9

Thomas Seeley wrote a nice book in which honeybee foraging behaviour, and the behaviour on the dance floor, is described in great detail:

• Seeley, T. D. (1995) The Wisdom of the Hive (Harvard University Press). ISBN 9780674953765. pdf version