Vince Gallo is a computer expert and keen beekeeper. When he discovered that scientists don’t truly understand how bees construct their perfect honeycomb, he joined Prof. Lars Chittka’s lab as a PhD student to find out. This is the first of a series of experiments with which Vince added more to our knowledge of honeybee architecture than has been done for a generation.

 

Honeybees building comb seem to work haphazardly, with different individuals coming and going, adding a piece of wax here, removing a piece there, often altering or undoing the work of other bees. There is no evidence to suggest that they are organised into teams and there is certainly no foreman or architect directing their efforts and yet from this seeming chaos emerges one of the most regular, geometrically perfect constructions to be found in nature. Vince’s key insight is that this process is managed by stigmergy, a form of self-organisational process in which the state of the workpiece itself guides each subsequent step. In other words, when a bee discovers certain features in the wax of the comb, they stimulate behaviours – likely following a small number of fairly simple rules – that change the state of the comb. The new shape of comb triggers a new set of actions (either in the same bee or a new one that stumbles across the partially built comb) which modify it still further, until a complete honeycomb has been built (and the process doesn’t even stop then, since comb must be constantly maintained and is often modified to suit the changing need of the colony, all probably directed through stigmergic processes).

 

In this paper, Vince examined how honeycomb is first started and then extended, cell by cell, but with each hexagonal cell neatly adjoined to its neighbours to tile the entire comb. He inserted small sheets of wax with various features into a hive and waited for the bees to construct honeycomb on them. Vince developed special software allowing him to precisely compare photographs of the stimuli with the resulting comb, making it clear how the final state of the comb was influenced by its starting conditions.

 

When presented with small indentations pressed into flat sheets of wax, it appears that bees start to deposit wax around the rim, beginning the process of building the walls that surround each hexagonal cell. They then extend the original depression, enlarging it to the required cell size. The wall of the eventual cell coincides with the original depression to a greater extent than expected by chance. When the initial wax had two indentations, close to one another, the bees built a pair of cells separated by a common wall which ran neatly between the pits, demonstrating that the choice of where to begin building a cell affects not only the position and spacing of cells, but also their orientation.

 

A v-shaped piece of wax attached to the starting stimulus slightly resembles the corner of a finished cell and Vince showed that that was enough to cause bees to build two new cells either side of the v, with their common wall extending from its point, to form the triple junction that defines the corners of all cells in a hexagonal grid. However, when the starting stimulus had both v-shapes and pits that were misaligned, the final construction had walls that were more closely aligned to the pits than to the v-shape, which makes it clear that it is the pits that trigger the building of each new cell which control wall placement, and not the corner of a previously built cell. In fact, it seems likely that when a bee encounters the outer wall of a cell rising from a flat base, that creates the same conditions as a depression and triggers the building of a new, adjoining cell.

 

Tiling the comb with hexagonal cells is the most efficient use of space, creating strong walls while using the minimum amount of wax. This work demonstrates how bees, working haphazardly, without any central organising force and almost certainly with no individual having a detailed plan or overview of the entire construction, can build such an impressive construction. A difference in height caused by a small depression, or a low wall can trigger the construction of a new cell wall, while the placement of individual cells, or even their starting conditions, causes new cells to be built in the right place to adjoin them. In truth, honeycomb is far from perfect – you can often see visible seams where separate piece of comb have merged, scars from previous damage, gradual changes in cell orientation, and odd spaces where cells don’t quite fit together – but the existence of simple rules that allow the current state of the comb to direct future construction explain how bees are flexible enough cope with these irregularities.

 

As well as completing his own PhD, Vince has been instrumental over the past few years in helping to develop the radars I use to track bee movements. Not only did he write a lot of the software I rely on, but he taught me a huge amount about computers, coding and, especially, technical project management. In return, I hope I was able to help him adapt his skills for academic research and teach him a little about experimental design, data analysis, scientific writing and so on. Other than that general advice over the course of the project, my contribution to this one was mainly in figuring out how to analyse the data and make it answer our questions.