Notes from Underground

Conventional wisdom about the ant world has long held that ants in colonies transition through age-related roles — from nurse to nest keeper to forager to soldier — throughout their lifetimes. But a College of Arts and Sciences biology professor has found evidence of brain changes called synaptic pruning as ants age, a surprising phenomenon that offers new insight into theories of ant social behavior. James Traniello (CAS’74) and his former graduate student Marc Seid (GRS’04) have found that instead of changing tasks, Pheidole dentata ants stack them: they begin as nurses, then progress to other tasks, but don’t stop nursing. “They add in nest maintenance and foraging, but never become incapable of going back to doing what they did when they were very young — and in fact, doing it better than they did it at first,” Traniello says. “There’s an enormous amount of behavioral flexibility.”

Such discoveries have come to Traniello with help from a nearly half-million dollar grant from the National Science Foundation. The neurobiologist and his students conduct their research in the woods in Concord, Mass., looking for living colonies of the ant species Pheidole pilifera. They excavate the nests, suck the ants up with an aspirator, then sort them out, making sure they have the queen, essential to keeping a colony alive. The purpose? To see what happens in ant brains that leads to behavioral specialization. These aren’t just idle questions: Edward O. Wilson, one of Traniello’s doctoral advisors at Harvard, has famously suggested that ant behavior is helpful in understanding the behavior of certain upright biped creatures as well.

Synaptic pruning, which allows ants to take on new tasks while continuing the old, seems related to neural plasticity, or the ability of multiple parts of the brain to accomplish the same task. In early human childhood, for instance, many different neural networks do the same work. But as we get older, networks are pruned back, leaving fewer but stronger connections. It had been thought that ants, since they have age-related tasks, have synaptic networks for each task, and thus no need for synaptic pruning. Traniello and Seid were the first to show that ants do in fact experience it. “Maybe it’s not surprising,” Traniello says. “Perhaps the same principle applies due to a common adaptive goal: lower the metabolic cost of maintaining the brain.”

It begs the question, then, as ants, with a life expectancy of about 80 days, mature — they are perhaps 5 days old when they first start working, maybe 20 when some start to go out and forage — are they actually learning tasks? “We think there is learning going on and that there is a neural plasticity here which we see at a number of different levels,” Traniello says. The plasticity is developmental and it’s functional: there are different castes, and each caste’s brain significantly differs from another’s.

Ants, like humans, are influenced by neurotransmitters called amines — dopamine and serotonin, for example, plus octopamine, which is exclusive to invertebrates. Traniello and his colleagues are examining how amines influence species traits, such as aggressiveness, and how they regulate brain development. “In dentata, at the time they leave the nest, you see a spike in serotonin and then a steady increase in dopamine.”

Amines basically modulate behavior. “The cool thing is that you can alter the levels of amines by feeding them to ants — just like you can take Prozac and alter the amine levels in your brain,” says Mario Muscedere (GRS’09), a doctoral student in Traniello’s lab. “We know serotonin levels increase with age. If we feed serotonin to young ants, will they start acting like old ants? Will they start going outside and foraging, as old ants do?”

These aren’t just hypothetical questions. “We’re going to do some Frankenstein kinds of experiments,” Traniello says, trying to manipulate behavior by changing amine levels. “The pilifera majors are very shy, so we assume they have an amine profile that’s different than morrisi, who are the most aggressive. So we’ll give the pilifera major the amine profile of a morrisi major — like you feed Prozac to people — and we should be able to create an aggressive pilifera major.” Very preliminary results from his lab seem to confirm these hypotheses, and more work is ongoing.

One of the most important elements of ant social evolution is division of labor, which theoretically leads to more efficient colony functioning, according to Traniello. “We feel we can illuminate the evolution of division of labor by understanding how the brain works and is structured in respect to task performance,” he says. “Brain structure reflects natural selection.”

By comparing how the brain works in the three species of Pheidole, which are closely related but differ in ecological settings and worker social roles, Traniello hopes to understand the evolution of division of labor, an important aspect of social insect life. “We can determine if a basic pattern remains constant across species, perhaps because it evolved a long time ago, independent of ecology,” he says, “or if adaptation has shaped each species’ social organization and division of labor differently.” He points to the parallel with understanding human evolution through development of the human brain. “Our scale, of course, is very different. But in both humans and ants, brains are instrumental in the establishment of social networks.”

Each summer, the life cycle of ant society begins again. Ant colonies release males and a few would-be queens to mate after rainfall in early July, and against the odds — only one in several hundred females who mate is successful in establishing a colony — start the life cycle all over again, as they have for 100 million years.

In a mere four years, a drop in that bucket, Traniello hopes to have at least the beginnings of an answer to his own questions about ant social evolution. “A lot of the work I did early on was trying to understand the evolution of caste and division of labor, which is really the heart of social insect organization,” he says. “To revisit the questions I started on 30 years ago, but with tools of neurobiology, will be very powerful. I think we’re going to make some real progress in understanding what goes on in social systems by studying brain evolution.”

Click here to read "Part One: BU researchers peer into ant brains to find out how they learn."

Taylor McNeil can be reached at tmcneil@bu.edu.