Study Shows Pinyon Jays Infer Their Status from Observing Others
Released on 08/24/2004, at 2:00 AM
Office of University Communications
University of Nebraska–Lincoln
Transitive inference is an ability that humans acquire early in life. It's a simple task for our minds to understand that if A is better than B and B is better than C, then A is obviously better than C, even if we've never seen A and C together.
Researchers have argued that this ability may reflect the evolution of humans as a highly social species, that transitive inference is one of the ways we figure out where we fit in society. Evidence from field studies of other social species, especially primates, suggests that they also use transitive inference in social settings. But such social inferences had never been clearly demonstrated in controlled experiments on non-human species, until a team of University of Nebraska-Lincoln-based scientists published their experimental results in the Aug. 12 issue of Nature, the international weekly journal of science.
The striking thing about how they demonstrated the use of transitive inference by another species, though, is that the scientists didn't use primates or even another species of mammal to prove it. They used birds -- pinyon jays (Gymnorhinus cyanocephalus), highly social members of the corvid (crow) family that live in permanent flocks of as many as 500 individuals.
"I don't think anybody will be surprised that this sort of behavior occurs, but I think a lot of people will be surprised that it occurs in birds," said Alan Bond, research associate professor of biological sciences at UNL and co-author of the Nature paper with Alan Kamil, university professor of biological sciences at UNL, Guillermo Paz-y-Mino, a former UNL post-doctoral fellow, and Russell Balda, emeritus professor of biology at Northern Arizona University in Flagstaff.
"This type of thing has been postulated in primates for years, although the evidence isn't terribly good," Bond said. "But I think if somebody went out and did it on baboons, people would say it's obvious. To show it in birds means something very different, though.
"There's been a lot of speculation in the literature for years about the origins of human intelligence and why it is that people have evolved to be able to do such complicated things with their brains. One of the big ideas that has emerged is what's called Machiavellian intelligence, and the idea is that a lot of the features of human cognition evolved in the context of human societies, as ways of dealing with the complexity of human social systems. We were able to show pretty clearly for the first time a cognitive adaptation, a way the brain is structured, that is related to how the social system works."
Bond and Kamil, who is also a professor of psychology, have studied avian cognition for years, often teaming with Balda. From their previous studies, they figured there was no reason why highly social species of any kind couldn't use transitive inference.
"One of the things that you need to do in a big, complex society is you need to know your place and your place relative to everybody else's," Bond said. "In a social group of pinyon jays as large as 500 individuals, it's nearly impossible to imagine that these birds work it out one-on-one. There would be a huge advantage to being able to make use of the interactions between other individuals, to figure out what your place in the society was."
The scientists first established that pinyon jays use transitive inference through a relatively simple laboratory study that asked pinyon jays and scrub jays (a closely related but nonsocial species) to learn the correct order of colored stimuli. The pinyon jays had very little trouble with the task, learning at the level of some monkeys in similar tests. The scrub jays flunked, performing no better than pigeons.
Showing that pinyon jays use transitive inference in social settings was more complicated, however, because many more factors had to be controlled.
The scientists captured 16 adult male pinyon jays that Balda had banded and studied in the wild in northern Arizona, housed them individually, and kept them mildly hungry (but not to the level of hunger that the birds often experience in the wild). The birds were divided into three groups (two of six birds each, one of four) so that no birds of separate groups had been in direct contact in the wild for at least five years.
In a laboratory in UNL's Manter Hall, the scientists set up a transparent acrylic chamber that measured roughly 40 inches long by 16 inches wide and 16 inches high and divided it into three compartments with sliding opaque and transparent dividers. They staged five-minute encounters between two birds within each group in which one bird was placed in each of the end compartments and the central compartment was baited with a single unshelled peanut. The birds were released simultaneously to compete for the peanut, and the scientists recorded which birds displayed dominant or submissive behaviors. (Who captured the peanut wasn't counted, because sometimes the submissive bird was clever enough to get it away from the dominant bird.)
The intragroup encounters enabled the scientists to learn the linear dominance hierarchy of the birds in each group. Once they knew the hierarchies, they began staging intergroup encounters, leaving out the top and bottom birds from each group because they were not accustomed to both winning and losing.
For the intergroup encounters, however, a third bird observed the encounter from outside the acrylic chamber. In experimental conditions, the observer bird watched a bird from his group against a stranger from a second group. For example, in Group 1, D lost to B. He watches B win and lose encounters against demonstrators from other groups. D is then tested against the demonstrator who dominated B. In control conditions, D watched an encounter between two strangers from the other groups, then was tested against the dominant bird.
The results under experimental conditions were dramatic.
"When we pair a bird with an individual that he has seen beat someone that the observer knows is dominant to him, he behaves submissively as soon as you open the doors," Bond said. "He sees this guy and says, 'I know that I'm going to have trouble with him,' and he starts saying, 'Oh, master, master.' The birds have about six different displays for showing submissive behavior and he starts that right off.
"But in the control situation, he doesn't know anything about any of the birds that he has seen, and there's basically no effect. What's really clear here is that these birds form expectations based on what they know about the birds they see interacting with other individuals."
CONTACT: Alan Bond, Research Assoc. Professor, Biological Sciences, (402) 472-0963