EXPERIMENTS ON BEES DONE BY JACK DARLEY AND BERTRAND ROEHNER AT MEUDON VAL FLEURY (WEST OF PARIS, FRANCE) IN JUNE-JULY 2012 For the first three experiments described below the positions of all bees in the course of time were recorded and stored into a computer. Average inter-bee distances were computed. This analysis confirmed the robustness of the evolution leading to the formation of a cluster. Most of the other experiments were rather preliminary in the sense that we tried to define the conditions under which specific effects (e.g. bees converging toward a small box containing other bees) could be observed. --------------------------------------------------- 1) Experiment of 13 June 2012, 14:30-15:40 This was a repeat experiment of the basic experiment described by Lecomte in his article of 1950 and which had already been repeated in Kunming in late November 2011. The experiment was performed outside at a temperature of about 20 degrees. Between taking successive pictures the box was closed so that the bees remained in the dark. There were 294 bees which had been taken from the frame of a beehive. Initially, they were put ``asleep'' with carbon dioxide. The surface of the box was 24.5 cm times 30.5 cm (its depth was about 30 cm) which means that on average each bee had an area corresponding to a square with a side length of 1.6 cm; this gives the average initial distance between bees. After about one hour, two clusters had formed on opposite sides. Taking advantage of one of the openings of the box, the cluster on the right-hand side left the box and flew away leaving a single cluster. --------------------------------------------------- 2) Experiment of 15 June 2012, 14:40-16:10 This was basically the same experiment as the previous one with however a few improvements. The box was larger: 39 cm times 50 cm. As there were 305 bees, each of them had a square area with a side length of 2.4 cm. The depth of the box was only 5.5 cm and it was permanently covered with a transparent plactic sheet (PVC or polyvinyl chloride, 0.5 mm thick). In addition an opaque cover was put in place between the moments when the pictures were taken. The temperature was 20 degrees. After about one hour the bees had formed a cluster on the right-hand side which comprised 83% of them. This situation remained unchanged for 20 mn, then within 5 mn the cluster moved from the right- to left-hand side. The number of bees which were not included in the cluster remained almost the same at a level of 15%. --------------------------------------------------- 3a) Experiment of 20 June 2012, part a, 14:30-14:45 After having been put to sleep the bees were positioned at the nodes of a square lattice. The spacing between two bees was 8 cm. There were 24 bees altogether. The temperature was 24 degrees. The idea behind this experiment was to observe what the bees would do after waking up. How quickly and how much would they come closer to one another? A picture was taken every 3 minutes but this interval turned out to be too long; indeed, even after 6 mn the locations of the bees were completely changed. Not unexpectedly, what was observed is that within 15 mn the bees formed a few small clusters. --------------------------------------------------- 3b) Experiment of 20 June 2012, part b, 15:00-15:40 The small box in the middle contained about 30 bees. The box was closed by nylon stockings on all its faces except one. Aside from the box there were 167 bees. Both the bees in the little box and those outside had been put to sleep before the start of the experiment. The idea was to see whether or not these bees will be attracted by the bees inside of the little box. It turned out that they were not. Within 40 mn they formed a cluster which comprised about 90% of the 163 bees. However, with only a few individual exceptions, the bees did not move toward those in the little box. --------------------------------------------------- 4) Experiment of 13 July 2012,14:40-15:25 This experiment was a second attempt to repeat the observation decribed in Lecomte's paper of 1950 in which the bees are attracted by those contained in a small box. With respect to the previous attempt, several changes had been made. * Instead of only 30 bees there were 270 bees in the small box * The bees in the small box were not put to sleep before being introduced. * Outside there were two small groups of bees: one at the center-bottom had 30 bees while the one on the left-hand side had 25 bees. Under such conditions about 94% of the bees clustered on the small box within 45 mn. The temperature was 18 degrees. --------------------------------------------------- 5) Experiment of 18 July 2012, 14:30-16:05 This is a continuation of the previous explorations in slightly different conditions. * There were 105 bees in the group on the left-hand side (which were put to sleep before the experiment) and 120 bees in the small box (not put to sleep). In fact, we wanted more bees in the small box but in the process of filling it a number of them flew away. * The experiment took place in a box which had the following inside dimensions: 93 cm times 17 cm and a depth of 7.5 cm. It was closed with a PVC sheet in the same way as previously. Except during the short time intervals for taking the pictures, the box remained in dark. The temperature was 25 degrees. Under these conditions about 50% of the bees moved toward the small box within about one hour. However, most of the bees which were around the small box did not cluster ON the box but rather on the wooden wall of the big box behind the small box. This was however a fairly temporary situation. In the following 30 mn almost all the bees which were near the small box moved back to their initial position on the left-hand side. --------------------------------------------------- 6) Experiment of 20 July 2012, 14:30-15:30 As the previous experiment did not give any clear result, we repeated it by making the conditions more similar to those in experiment 4. * There were some 200 bees in the small box. * Outside there were two groups each of 30 bees. The temperature was 21 degrees. Under these conditions, almost all 60 bees clustered on the small box within 25 minutes. However, they did not remain on the box. About 20 minutes later only a few were still on the box. --------------------------------------------------- 7a) Experiment of 25 July 2012, part a, 14:20-15:09 This is what can be called a two-compartment experiment. The container is divided into two parts of same area that we call part 1 and part 2. There are two possible initial conditions. * Bees can be put in equal numbers in part 1 and in part 2 and one observes the fluctuations of these numbers. * Bees can be put in part 1 while part 2 is left empty. In the present experiment we tried the second method. The objective was to record the number of bees which left part 1 and to see whether an equilibrium is reached after some time. This equilibrium would then be similar to the one observed with a liquid (in part 1) on top of which there would be molecules of vapor which escaped from the liquid. This experiment produced quite unexpected observations. The weather was fairly hot with temperature reaching 35 degrees. In this and the following experiment the box was not darkened between the picture shootings but as it was in the shadow of a tree the bees were not exposed to sunlight. At the start, after being poured into compartment 1, some 5% of the 212 bees crossed into part 2. This can be attributed to the ``excitation'' brought about by such a rather rude treatment. It can be noted that at that moment many bees had their wings perpendicular to their body. After about 5 mn only some 2% remained in part 2. Some 25 minutes after the start, almost no bees remained in part 2. Then, suddenly, the bees started to fan their wings (which produced a roar sound) and within 2 minutes some 27% of them moved into compartment 2. as previously this situation did not last and 10 minutes later only few remained in part 2. In conclusion one can say that it was difficult to observe an equilibrium situation in which some of the bees were in part 2. The only lasting condition was with almost all bees in part 1. However, this condition was sometimes disturbed by a sudden push of bees toward the other compartment. It is possible that the high temperature played a role in this behavior (especially with respect to fanning with the wings). --------------------------------------------------- 7b) Experiment of 25 July 2012, part b, 15:23-16:22 This experiment was the same as the previous one except that the bees were initially put to sleep before being put in part 1. The result was that within one hour nothing happened. All 68 bees remained in part 1. The only evolution was a slow diffusion process which made the initial cluster looser. --------------------------------------------------- 8) Experiment of 27 July 2012, 14:48-15:53 This experiment was similar to experiment 5 and it led to the same inconclusive results. There were some 100 bees in the small box and some 50 bees outside. Throughout the experiment there were never more than 20% of the bees clustered on the small box. --------------------------------------------------- CONCLUSIONS The objective of this series of experiments was to define a method for measuring the interactions in a group of bees. Four methods were tried: (i) Cluster formation (ii) Bees on a lattice (iii) Attraction by a group of bees in a box (iv) Two-compartment experiment All these methods converged in showing that there is a strong attraction between bees. Cluster formation seems very robust and reproducible; unfortunately, this method cannot be used with living organisms which do not form clusters. Attraction by bees in a small box is also fairly reproducible provided that there are enough (over 200) bees in the box. At this point we do not know whether this method also applies to living organisms which have weaker interactions. The lattice bee experiment seems promising but must be tried further. The two-compartment experiment seems promising for living organisms whose attraction is weaker than for bees. For bees the attraction is so strong that the situation is that of liquid at low temperature (or even may be a solid) for which the equilibrium vapor pressure is so low that it is difficult to measure.