CLUSTERING EXPERIMENTS

Suggestion After clicking on one of the following sections, once the images are downloaded (which make take 2 or 3 minutes), it is necessary to zoom in strongly (Control+wheel of the mouse) on each image for otherwise the individual organisms will hardly be visible.

Clustering consists in the fact that, without any external incentive, scattered individuals gather closely together thereby forming an entity of much higher density. Among collective motions clustering is probably the simplest and arguably the most important. Why?
* Clustering can occur only if there is an attraction between individuals. If this attraction increases for shorter distances then, by coming closer together, the separate organisms are able to strengthen the interaction between them. This self-amplifying process is subject to the sole limitation that individual organisms must remain separate. As can be seen in the pictures given below, the resulting cluster is a stable and mostly static system.
In short, clustering is a direct consequence of inter-individual attraction and, contrary to many other collective motions, it does not involve elaborate dynamic behavior.
* The fact that the density of the cluster is much higher (typically hundreds of times higher) than the density of the initial population often gives it special qualitative properties. As a simple illustration one can mention the ``clustering'' of the molecules of a gas into droplets of water. Clearly, the physical properties of water greatly differ from those of a gas. The same is true for many other clusters. Cities have characteristics which do not exist in low density countryside habitat.

A. Bees

Apis mellifera (June 2012)

B. Small ants

Small ants (2 observations in parallel)

Small ants (4 observations in parallel)

Small ants (8 observations in parallel)

C. Big ants
These big ants (Camponotus Japonicus) walk much faster than the smaller ants used in the previous experiments. For that reason one would not expect them to form a cluster so easily. This is indeed confirmed by observation although eventually they did form a cluster. Actually, their behavior is fairly similar to the one described by Lecomte (see the papers above) for the case of bees: (i) clustering will only occur above a density threshold (ii) clustering will take longer than for small ants.

Big ants

D. Drosophilas
The question is: do drosophila form clusters?
The two experiments which follow try to provide an answer.

Drosophilas (1st experiment)
No clustering was observed but this experiment had to be cancelled because when it was finished we found out that the source of light located under the container raised its temperature by some 10 degres. Thus the drosophilas were not in normal conditions (they moved much faster than at room temperature). The experiment was repeated (see below) with a ``cold'' light-source.

Drosophilas (2nd experiment, 2 observations in parallel)
It turns out that the drosophila used in this experiment do not form any cluster even for densities much higher than those used previously for ants. In other words, it seems that drosophila are solitary insects.
Whether or not this can be due to the fact that, generations after generations, they were raised in laboratory conditions (as opposed to natural conditions) remains an open question.

We would like to express our gratitude to Pr. Dou Fei and Dr. Liu Xia from the Genetic Research Institute of Beijing Normal University for their kind help in this experiment.

E. Beetles
FILM: Clustering of beetles
The small beetles are ``Alphitobius diaperinus'', whereas the larger are ``Tenebrio molitor''. Both species are worms in their pre-adult stage and are sold in stores particularly to feed fishes and birds. The worm of the small beetle is also called ``buffalo worm'' or ``lesser mealworm''.
The present experiment lasted 5mn and a picture was taken every 5 seconds. The diameter of the cup was 15.5 cm. The experiment was performed in the evening (19:45) under fairly uniform artificial light. As can be seen, the beetles begin to move toward the edge of the cup. The same reaction is also observed for ants. Once the beetles are along the edge, their movements become more or less one-dimensional.
The experiment was performed on 5 October 2012 at the Bee Institute in Kunming (Yunnan province, China). Many thanks to Prof. Ken Tan and to his group. F. Euglenas
The Euglenas are unicellular protists which swim in water. Their length is about 50 micrometers. Similarly to plants, they contain pigments of chlorophyll. Highly photo-sensitive, they are attracted by low intensity light but above a given intensity threshold light has a repulsive effect.
When a solution containing Euglenas is subject to strong light, a clustering process occurs within a few minutes which produces a pattern of grey lines. Microscopic observation shows a high density of Euglenas around these grey lines (see the photos).
Observation of clustering Euglenas (pictures)

We would like to express our gratitude to Pr. N.J. Suematsu of the Department of ``Mathematics and Applications'' of Meiji University in Tokyo, to Pr. Dou Fei and Dr. Liu Xia from the Genetic Research Institute of Beijing Normal University for their kind help in this experiment.

FILM 1: Observation of Euglenas outside of a clustering area, magnification is x90 (format: mov)
FILM 1: Same film as above (format: gif)

FILM 2: Observation of Euglenas outside of a clustering area, magnification is x225 (format: mov)
FILM 2: Same film as above (format: gif)

For his kind help we would like to express our gratitude to Pr. Michel Gho of the ``Cell cycle and cell determination'' group (Laboratory of Development biology, University of Paris, Pierre and Marie Curie).