Insect Swarms May Resemble Star Clusters

Scientists have long been fascinated by the collective behavior of animals. The coordinated motion of schools of fish or flocks of birds, for example, can be built up from short-range interactions where individuals tend to move much like their neighbours. Swarming insects, however, do not align themselves in this way, and a different approach is required to understand the swarm’s dynamics. A team of researchers based the Weizmann Institute of Science and Stanford University has found a mathematical resemblance between insect swarm dynamics and gravitational interactions. The study, which has just been published in the New Journal of Physics, could represent a leap forward in understanding the mass movement of flying insects.

The scientists, led by the Weizmann Institute’s Prof. Nir Gov and Stanford’s Prof. Nick Ouellette, focused on midges which, in common with other insects such as fruit flies, moths and bees, have an organ on their antennae known as the Johnston's organ that is sensitive to sound. To the researchers, this highlighted the likely importance of acoustic interactions in swarming, where sound is created as the flying insects beat their wings. “We started from the assumption that when one midge hears another, it reacts by accelerating towards the source with a strength that is proportional to the perceived sound intensity,” said Dr. Dan Gorbonos, a postdoctoral researcher in Gov’s group in the Weizmann Institute of Science. “The decay rate of the sound intensity falls off according to an inverse-square law – just as the gravitational attraction between two masses does.”

In applying their model of collective animal behavior, the scientists found a similarity in the structure of midge swarms and astrophysical objects such as globular clusters that consist of hundreds of thousands of stars. However, there is a major difference between the real gravity that acts between masses and this effective acoustic interaction that the group dubs “adaptive gravity.”

What is adaptive gravity? To understand the team’s approach, it is helpful to consider that for many animals the perception of sound is not fixed, but instead adapts to the total sound intensity. Our ears, for example, behave in this way. We find it easy to decipher a whisper in a quiet environment, but people have to shout to be understood in a noisy room. “We expect to find a similar mechanism in the Johnston's organ,” commented Gorbonos. “Therefore, we would expect that the midges will reduce their sensitivity in a densely populated environment, and we found evidence for this in the experimental data.”

This adaptivity feature helps to explain why the swarm doesn’t collapse in on itself despite the gravity-like interaction. When the background buzzing noise is too high, and the local environment is too dense, the sensitivity is lowered and each individual midge reacts less to the attractive pull of its neighbours.

The scientists filmed the midges and analysed their movements using a system that was developed by Ouellette, and Drs. James Puckett and Rui Ni of Yale University. The software was originally developed for tracking turbulent flows in physics. The collaboration began when Oullette visited the Weizmann Institute of Science and described his midge research. His research caught Gov’s interest, and he suggested borrowing tools from gravitation to model the midge behavior. Dr. Reuven Ianconescu joined the team led by Gov to work on the simulations, followed by Gorbonos, whose background is in the area of gravitational physics.

The group sees this adaptive model as a prototype for descriptions of various kinds of collective systems with long range interactions – ranging from other acoustic interactions in the animal kingdom to much smaller scale biological systems, possibly even swarming cells.

Prof. Nir Gov's research is supported by the Yeda-Sela Center for Basic Research; and the Clore Center for Biological Physics. Prof. Gov is the incumbent of the Lee and William Abramowitz Professorial Chair of  Biophysics.