Sheep may come across as simple, straightforward followers. But as those who work with the animals know, sheep can behave in surprising ways. For one, flocking may not show a friendly tendency—there might be selfish motivations as well.

In research published on July 24 in the print edition of Current Biology, scientists monitoring sheep have demonstrated empirical support for a behavioral theory first proposed almost 40 years ago. "We're getting to the nitty-gritty of the real evolutionary rules selected for in behavior," says behavioral ecologist Andrew King of the University of London Royal Veterinary College, a co-author on the study. King has studied sociality in birds, fish, baboons and even humans—and believes these rules of group behavior found in sheep could apply more broadly. "If we can understand how these rules work, we can hint at similar systems and common principles across species."

Selfish-herd theory, first proposed by the British evolutionary biologist W. D. Hamilton in 1973, posits that individuals dilute their risk of predation by moving into a larger group—a "Don't eat me, eat this other guy!" response. Although widely cited, the theory is difficult to test in the field given the unpredictable nature of predators and challenge of observing movements in a large group of animals.

The interactions between sheep and sheepdog, however, provide an ideal testing ground. "Working with sheep is great because you can measure their movements and even tell the dog when to begin," King says. "You have control of the whole system but it remains really natural as well."

King—with colleagues at the University of London and University of Cambridge—outfitted 46 sheep with GPS backpacks to enable second-by-second remote tracking. They then orchestrated a predator event by releasing an Australian Kelpie, a sheepdog that also wore a GPS module, on the flock. In each of three trials, the approaching dog drove the sheep in a fleecy flurry while scientists recorded how each individual in the group moved.

In each trial, the sheep ran rapidly toward the center, jostling for position within the group, as selfish-herd theory would predict. To further suss out the rules behind the sheep's movements, the biologists compared the observed behavior with various mathematically modeled scenarios. As they report in Current Biology, one of the patterns they noted relates to the kind of predator behavior that triggers a sheep's flight. Whereas one hypothesis suggests an individual will start running when a predator is within a given distance, the researchers found instead that sheep won't start moving until the dog has approached a relative midpoint of neighboring sheep. The finding suggests that sheep have some level of social-spatial awareness, possibly because they do not want their herd scattered.

This particular predator–prey model may have a few confounding elements, however, particularly because the sheepdog has no intention of eating the sheep and the latter may be familiar with the dog's approach. Nevertheless, the authors found the response of the sheep to be so consistent that they believe it makes for a reliable representation of predator–prey interaction.

"How individuals move relative to one another within groups to avoid predation has remained an intriguing yet vexing field of research," says behavioral ecologist M. Justin O'Riain at the University of Cape Town in South Africa. O'Riain, who has studied how seals move into more compact groups when a shark approaches, notes that the innovative use of GPS technology in the study by King and colleagues takes these questions to "a new and exciting level."

Lesley Morrell, an evolutionary biologist at the University of Hull in England also not associated with the study, observes that this kind of behavioral modeling may allow us to answer many questions. "There is a huge gap in our knowledge here in that we don't know which, if any, of the movement rules that have been proposed actually operate in real animals," Morrell says.

In addition, co-author Jenny Morton, a neurobiologist at Cambridge, hopes to incorporate this work in her efforts to understand the full repertoire of normal sheep behavior. Morton believes that sheep—with their large brains, developed cortices and relatively long life spans (about 15 years)—may prove stronger model organisms for studying neurodegenerative disease than either rodents or monkeys. She is currently investigating a sheep model of Huntington's and Batten's diseases. By fully detailing sheep social behavior, she hopes to recognize anomalous behavior in diseased sheep as well as track changes over the course of therapeutic treatments.

She adds that her study of the animals has been a continual source of surprises: "They're actually very clever in a 'sheepy' kind of way," Morton says. "They're not going to put a sheep on the moon, but sheep do remember faces, they recognize people and have long memories for complicated things. They're quite curious creatures."