200972211.jpg

Stephen Pratt/Arizona State University

Making sense. An ant colony scurries around its new home--the result of a rational decision.

Can't Decide? Ask an Ant

By: 
Michael Torrice
2009-07-22 (All day)

Classical philosophers called humans "the rational animal." Clearly, they never looked closely at ants. A new study suggests that ant colonies avoid irrational decisions that people and other animals often make.

Consider the following scenario: You want to buy a house with a big kitchen and a big yard, but there are only two homes on the market--one with a big kitchen and a small yard and the other with a small kitchen and a big yard. Studies show you'd be about 50% likely to choose either house--and either one would be a rational choice. But now, a new home comes on the market, this one with a large kitchen and no yard. This time, studies show, you'll make an irrational decision: Even though nothing has changed with the first two houses, you'll now favor the house with the big kitchen and small yard over the one with the small kitchen and big yard. Overall, scientists have found, people and other animals will often change their original preferences when presented with a third choice.

Not so with ants. These insects also shop for homes but not quite in the way that humans do. Solitary worker ants spread out, looking for two main features: a small entrance and a dark cavity. If an ant finds an outstanding hole--such as the inside of an acorn or a rock crevice--it recruits another scout to check it out. As more scouts like the site, the number of workers in the new hole grows. Once the crowd reaches a critical mass, the ants race back to the old nest and start carrying the queen and larvae to move the entire colony.

To test ant rationality, Stephen Pratt, a behavioral ecologist at Arizona State University in Tempe, and a colleague designed a series of possible nests for 26 ant colonies. The duo cut rectangular holes in balsa wood and covered them with glass microscope slides. The researchers then drilled holes of various sizes into the glass slides and slipped plastic light filters under the glass to vary the features ants care about most. At first, the colonies only had two options, A and B. A was dark but had a large opening, whereas B was bright with a small opening. As with humans, the ants preferred both options equally: The researchers found no difference between the number of colonies that picked A versus B.

This item requires the Flash plug-in (version 8 or higher). JavaScript must also be enabled in your browser.

Please download the latest version of the free Flash plug-in.

Follow me! A scout recruits another ant and leads it to a possible new home.
Credit: Stephen Pratt/Arizona State University

Then the scientists added a third option, called a decoy, that was similar to either A or B in one characteristic but clearly worse than both in the other (a very bright nest with a small opening, for example). Unlike humans, the ants were not tricked by the decoy, the team reports online today in the Proceedings of the Royal Society B. Although a few colonies picked the third nest, the other colonies did not start favoring A or B and still split evenly between the two.

Pratt speculates that ant colonies avoid making the irrational decision because, unlike humans, each ant doesn't evaluate all options before making a choice. When the scouts find a nest, they're unaware of what else is out there, and either they pick the nest or they don't. "The group may do better precisely because the individuals are ignorant," Pratt says.

Melissa Bateson, an ethologist at Newcastle University in the United Kingdom, believes the findings could have "really interesting implications for the benefits of collective decision-making, which we could learn something from." Many grant-review boards, for example, behave like ants: Reviewers see only a subset of the total grant applications and thus have to make a decision without seeing every option.

Go to Comments

Related Stories

What's New

Popular

  • Most Read
  • Most Commented
  • Past:
  • 1 week