“In the long history of humankind (and animal kind, too) those who learned to collaborate and improvise most effectively have prevailed,” wrote Darwin. Ants, a family that has inhabited the earth for about 100 million years, must be one of the most magnificent manifestations of such biological cooperation. Thousands of female ants pull together in a coordinated effort to ensure that all the needs are met for the proper functioning of the entire colony. (The males’ only role is to mate with the queen, and once this is completed, they die). While some ants forage, others stay behind to tend the brood, or to build, maintain or defend the colony’s living quarters; and there are even those whose task it is to bury the dead. But the thing that makes their behavior so remarkable is that they have no leader – no boss or governing body to allocate and manage their activities. How exactly do ants collaborate and divide the labor among themselves so successfully and altruistically?
In his new lab in the Physics of Complex Systems Department,
Dr. Ofer Feinerman and his team are hoping to reveal some of the ants’ secrets in a collaborative effort of their own, using tools from such fields as information theory, statistical and theoretical physics, computer science, systems biology, neuroscience and, of course biology. “Biology is rife with complex systems consisting of individual components – proteins, cells, organisms – organizing themselves into networks to coordinate their activity. While biology is able to identify and describe the individual components, the interactions between them can get very messy and analyzing such data can become overwhelming. By borrowing tools from physics and math, more quantitative measurements can be used to discern the rules that govern such complex collective behaviors,” says Feinerman.
Social networking
Ants primarily “speak” to one another in the language of chemicals: If an ant finds a rich food source, for example, it will deposit a trail of pheromones that tells the other ants where to find it. Social networks are formed between the ants during such communication, and it is these networks that Feinerman wants to understand.
In a setup reminiscent of Big Brother, Feinerman has handpicked a number of native Israeli ants to enter an artificial, nest-like structure that has cameras dotted around, enabling his team to eavesdrop on the ants’ “conversations.” Each ant is identified by a barcode glued to its back, enabling the scientists to track and record its activities. The scientists are hoping to answer such questions as: Who “speaks” to whom? Do ants form cliques and only interact with those in the same group, or are they indiscriminate? Do they employ far-reaching social networking tools akin to, say, Twitter or Facebook, relying on others to “retweet” and “share” their messages? Do messages get passed down a line of ants, or would an ant prefer to wander far and wide to make sure the message is relayed accurately?
Feinerman: “Ants use different communication strategies, ranging from one extreme to the other, depending on the setting. For example, in ‘piggybacking,’ one ant rides on top of another; though slow, this is a very reliable and direct method of communication, appropriate, for example, for ensuring the second ant arrives at the exact food location without getting lost along the way. Other situations may warrant less focused, but faster, forms of communication. For example, if a colony comes under attack, the ants spray pheromones into the air. This alarm system rapidly warns other, distant colonies of potentially imminent danger.”
Slow and reliable or fast with some compromise: Understanding ant communication and social networking strategies will do more than help reveal the intricate workings of an ant colony; they also assist in the long term goal of using this rich cooperative system to develop a theory of collective information processing. Ultimately, the researchers would like to develop the tools needed to answer fundamental questions about other complex biological systems: for instance, how immune system cells work together to fight infection. Such tools could also have practical applications in the design of so-called distributed systems, including cellular communication antennae, wireless sensor networks or even groups of robots engaged in rescue operations.
Empire of the Ants
At the head of each ant colony is a queen; but, though “royalty,” she does not possess any sovereign power over the worker ants – her sole duty is to lay eggs. With no ruler, how do ants divide labor? Do some ants belong to “elite” units, with others relegated to more “lower-class” duties? Or are they all equal?
In previous research, Dr. Feinerman investigated whether certain factors – previous experience, age, body weight, spatial location – determine how tasks are allocated. One way to test this is to see how ants respond to increased demands in certain tasks. For example, when food stores are running low, which ants take it upon themselves to aid in the foraging effort? Or offer a helping hand to nurse an extra brood in the colony?
The ants were placed in an artificial, two-chambered nest, and the scientists manipulated the task load by withholding food or adding new members to a brood. Again, the ants were individually tagged – this time using radio-frequency identification (RFID) – to identify individuals that responded.
This study revealed that it is the leaner ants that are usually on the frontline: They are the first to respond to increased foraging needs, even when other factors (age, experience and spatial location) are taken into account. They also seem to be the ones who engage in transporting the newly planted brood members to the main brood pile. However, lean ants do not seem to actually help in tending the brood. In fact, Feinerman found that brood care seems to be random, not dependent on by age, experience or weight, but rather, by whoever happens to be passing by.
“With regard to foraging, sending out the leaner ants could boost colony survival, as they would be more expendable in the risky task of foraging, may attract fewer predators, and might even be more mobile than their heavier counterparts. Likewise, the fact that brood care seems to be open to all suggests that flexibility in general task allocation, and collaboration on the whole, is more valuable than, say, expertise or experience, as it allows for a rapid response to a changing environment, thereby ensuring the survival of the colony,” says Feinerman.
ANT-thropologist
Dr. Ofer Feinerman was born in Rehovot, Israel, and he earned a B.Sc. in physics and mathematics
summa cum laude (1996) and an M.Sc. in physics (1999), both from the Hebrew University of Jerusalem. As a Ph.D. student at the Weizmann Institute of Science, under the guidance of
Prof. Elisha Moses, he grew the first artificial logic circuits made of nerve cells, earning his doctorate in 2006. He then conducted postdoctoral research for three years at the Memorial Sloan-Kettering Cancer Center in New York, investigating how the immune system's T-regulation cells work together to fight infection. Feinerman spent one year studying ants at Rockefeller University before joining the Department of Physics of Complex Systems at the Weizmann Institute of Science as a Senior Scientist in October 2010. “Coming back to the Weizmann Institute was a dream – not only did it give me the opportunity to return to Israel, but it also gave me the freedom to study ants – a somewhat unconventional subject. For this I am eternally grateful to and proud of the Weizmann Institute.”
Feinerman is married to Micka, a mosaic artist, and has three children: Matan (9), Shai (7) and Nomi (4).
Dr. Ofer Feinerman's research is supported by the Clore Foundation.
Materials on the Edge
In another part of the physics buildings, research students Oded Zilberberg in the group of Prof. Yuval Gefen, Yaacov Kraus in the group of Prof. Adi Stern and Zohar Ringel in the group of Prof. Ehud Altman were becoming interested in the new field of topological materials. These materials, discovered in 2007, are mostly normal, periodic crystals, but they have unusual electrical properties: The interior of the material acts as an insulator, but its surface conducts electricity just as if it were coated with metal. Like the discovery of quasicrystals, these new materials forced scientists to rethink accepted views – in this case, that a material was either an insulator (in which electrons don’t move) or a conductor (in which electrons can easily flow). Since then, scientists have been searching for additional topological materials, both with a view to possible applications and to investigate their unique properties.
Working together, the five students succeeded in showing – in theoretical work and in experimental measurement – that quasicrystals do, indeed, have topological properties. In some ways these newly-discovered properties are similar to those of “regular” topological materials, but in other ways they are very different.
In experiments conducted in Silberberg’s lab, the students built a quasicrystal using an array of coupled thin optical fibers etched with a powerful laser beam into a single glass cube. The parallel fibers were arranged at quasiperiodic (non-repeating) distances. When light was shined into a central fiber in this system, it “hopped” to the other fibers, emerging from all the fibers at the other side. But when the light was introduced at one of the far edges of the array, it remained confined to that side. In other words, the quasicrystal setup showed an “edge state” with different properties than the states found in the middle of the array. Just as electrons in topological materials flow only on the surface and do not penetrate the interior, the light shined through the outer edge of the optical quasicrystal setup stayed in that plane. This finding was a surprise, as scientists had believed such topological behavior to be impossible in a one-dimensional system.
Adiabatic pumps are known from a different kind of topological system called a quantum Hall system, in which electrons are exposed to directional magnetic fields. The model the students developed to explain this behavior enabled them to envision the optical quasicrystal setup as a sort of one-dimensional cross section of a two-dimensional quantum Hall system. That cross section, to the team’s surprise, preserves the topological properties of the two-dimensional system. In other words, one-dimensional quasicrystals “inherit” their topological properties from higher-dimensional “ancestors,” which are periodic crystals.
These findings were recently published in Physical Review Letters, and they were highlighted in Science and other scientific journals. Now, among other things, the team is checking whether their results are applicable to other dimensions. For instance, the theory they developed suggests that topological properties in three-dimensional quasicrystals can be linked to topological systems existing in six dimensions.
Dr. Ehud Altman's research is supported by the Yeda-Sela Center for Basic Research; and the estate of Ernst and Anni Deutsch.