Symmetry breaking of escaping ants

From Wikipedia, the free encyclopedia
Jump to navigation Jump to search

Symmetry breaking of escaping ants is a phenomenon that happens when ants are constrained into a cell with two equivalent exits, and perturbed with an insect repellent. Contrary to intuition, ants tend to use one door more than the other on average (i.e., there is a symmetry breaking in the escape behavior), so they crowd on one of the doors, which decreases the evacuation efficiency.


The symmetry breaking phenomenon arises in experiments described as follows. Worker ants freshly collected from the field are enclosed into a circular cell with a glass cover in such a way that they can only move in two dimensions (i.e., ants cannot pass over each other). The cell has two exits located symmetrically relative to its center. We will describe first a "reference" experiment, and then the one where the "escape symmetry" is broken.

In the reference experiment, both doors are opened at the same time and let the ants escape. If the experiment is realized many times, in average we see that approximately the same number of ants use the left and the right doors. In this experiment the escape symmetry is not broken.

The second experiment involves a further step before opening the doors: an insect repellent fluid is poured into the cell at its center through a small hole in the glass cover. As a result, ants get very excited. If the experiment is realized many times, we see that the number of ants escaping by one of the doors (which can be randomly either the left one or the right ones) is significantly higher than the number of ants escaping by the other one, i.e., the escape symmetry is broken. The crowding of ants at one of the doors while the other one may be eventually free results in an inefficient evacuation in terms of time.

Geng Li and coworkers from Beijing Normal University tested whether and how the density of a group influenced symmetry breaking in escaping ants. They used the red imported fire ant (Solenopsis invicta) to repeat the experiment mentioned above with different total number of ants. The result shows that the symmetry breaking increases at low density but decreases after a peak. That is to say, when density is low, the ant group produces a collective escaping behavior; while at high density, the ant group behaves more like random particles.[1]


Inspired by earlier computer simulations that predicted a symmetry breaking phenomenon when panicked humans escape from a room with two equivalent exits, E. Altshuler and coworkers from the University of Havana designed the experiment described in the section above, which revealed the symmetry breaking effect in the leaf-cutting ant Atta insularis.[2]

The ant Atta insularis (commonly called bibijagua in Cuba) displays a high degree of swarm intelligence, like most social insects, which has made them great survivors through millions of years of evolution. In Cuba, a person may be described as "smarter than bibijaguas" ("sabe mas que las bibijaguas"). The poor evacuation efficiency illustrated by the phenomenon of symmetry breaking in escaping ants is one of the few examples where bibijaguas do not seem to behave collectively in a smart way.


The basic idea is that the action of the repellent induces herd behavior in the ants. When ants are in "panic", they experience a strong tendency to follow each other. As a result, if a random fluctuation in the system produces a locally large amount of ants trying to reach one of the two doors, the fluctuation can be amplified because ants tend to follow the direction of the majority of individuals. Then, that specific door becomes crowded.

E. Altshuler and coworkers were able to reproduce their symmetry breaking experiments in ants using a simplified version of the theoretical model proposed earlier by Helbing et al. for humans,[3] based on the fact that walkers tend to follow the general direction of motion of their neighbors ("Vicsek's rule"[4]), and such herd behavior increases as the so-called "panic parameter" increases. In the case of ants, the panic parameter is supposed to be low when no repellent is used, and high when the repellent is used.

A more "biologically-sensible" model based on the deposition of an alarm pheromone by ants under stress also reproduces the symmetry breaking phenomenon, with the advantage that it also predicts the experimental output for different concentrations of ants in the cell.[1]

The pheromone mechanism shares the key element of the previous models: stressed ants tend to "follow the crowd".

Ants vs. humans escaping in panic[edit]

From the statistical point of view, human beings and ants seem to behave in a similar way in certain scenarios, such as evacuation from a room or cell under stress, resulting in symmetry breaking phenomena: while ants increase their tendency to follow each other under stress, human beings seem to forget about Emergency evacuation strategies, and just run towards the door where most people run to. However, the enormous physical and behavioural differences between humans and ants imply also big differences in their collective behavior.[5][6]

In the symmetry breaking experiment on ants, for example, the individuals are relatively "polite": rarely ants are crushed by other ants in the process or evacuation. When humans are in panic, however, deaths by overpressure or suffocation are common, like in the case of the Love Parade disaster occurred in Germany in 2010.

Videos on symmetry breaking in escaping ants are available on YouTube.[7]


  1. ^ a b Li, G.; Huan, D.; Roehner, B.; Xu, Y. J.; Zeng, L.; Di, Z.; Han, Z. G. (2014). "Symmetry Breaking on Density in Escaping Ants: Experiments and Alarm Pheromone Model". PLOS ONE. 9 (12): 0114517. Bibcode:2014PLoSO...9k4517L. doi:10.1371/journal.pone.0114517. PMC 4281238. PMID 25551611. open access publication – free to read
  2. ^ E. Altshuler; et al. (2005). "Symmetry breaking in escaping ants". American Naturalist. 166 (6): 643–649. doi:10.1086/498139. JSTOR 498139. PMID 16475081.
  3. ^ D. Helbing.; et al. (2000). "Simulating dynamical features of escape panic". Nature. 407 (6803): 487–90. doi:10.1038/35035023. PMID 11028994.
  4. ^ T. Vicsek; et al. (1995). "A new type of phase transition in a system of self-driven particles". Phys. Rev. Lett. 75 (6): 1226–1229. arXiv:cond-mat/0611743. Bibcode:1995PhRvL..75.1226V. doi:10.1103/PhysRevLett.75.1226. PMID 10060237.
  5. ^ C. Detrain and J.-L. Deneubourg (2006). "Self-organized structures in a super-organism: do ants "behave" like molecules?". Physics of Life Reviews. 3 (3): 162–187. Bibcode:2006PhLRv...3..162D. doi:10.1016/j.plrev.2006.07.001.
  6. ^ S. Boari; et al. (2014). "Efficient egress of escaping ants stressed with temperature". PLOS ONE. 8 (11): e81082. Bibcode:2013PLoSO...881082B. doi:10.1371/journal.pone.0081082. PMC 3843683. PMID 24312264.
  7. ^ Complexperiments (2012). "PANICAntsEng".