The wisdom of the swarm

The propensity of locusts to form huge swarms and blanket landscapes may have evolved as a strategy to disrupt foraging by predators such as small mammals, lizards, and birds, according to research published today.

"If the [locusts] are dispersed and in large numbers, a predator can move through an environment preying on the food and sustain itself," explained Andy Reynolds, a biomathematician at Rothamsted Research in the UK and lead author on the paper, which appears in today's (Dec. 18) issue of Current Biology. "If the [locusts] bunch up and form large groups, then from the point of view of a predator, the environment is sparse."

A locust swarm in Australia

Courtesy of Andy Reynolds

Reynolds, with colleagues in the UK and Australia, applied "percolation theory," a mathematical model commonly used by material scientists, chemists and physicists to describe the behavior and connectivity of random clusters of atoms or molecules, to explain the mass aggregations.

The researchers calculated a "threshold of percolation," which describes a critical level of connectivity between groups of locusts below which predators choose to give up on finding more prey items. When locusts are grouped in small clusters in a constrained geographic area, they constitute a connected patchwork, through which predators navigate, eating locusts as they go. As locust population density in that area increases due to favorable climatic conditions, the small clusters become larger and predators find it even easier to move between groups, eating as they go. Reynolds said that it is advantageous for the insects to switch from small group living to form swarms as the connectivity between groups nears the threshold because predators are less likely to home in on widely-separated, less connected mega-groups than they are on tightly connected small groups. Instead of moving from locust clump to locust clump in a connected patchwork, predators are faced with a giant clump that is more separated from neighboring swarms and is thus harder to locate and move between. "[The model] shows that there is a threshold below which a predator cannot sustain itself [by finding enough prey items] and above which it can," Reynolds said.

Previous models had suggested evolutionary advantages, such as predator satiation and decreased detectability, from the perspective of an individual locust flying in a swarm, but Reynolds's is the first to model the sharp transition from non-swarming to swarming lifestyles. "There's a sudden and abrupt change," Reynolds told The Scientist. "Other models have not seen this sudden change."

"It's a very neat and elegant mathematical model," Lancaster University evolutionary ecologist Kenneth Wilson told The Scientist. Wilson was not involved in the study but did write a companion piece to the study that also appears today in Current Biology.

Desert locusts exist in two phenotypic forms -- the green, cryptic form, where the insects generally avoid contact with conspecifics and the swarm form, where yellow and black locusts join together in huge swarms with voracious appetites. The switching from one form to the other is one of nature's most striking examples of "phase polyphenism."

A solitary, green locust with a gregarious yellow and black one.

Courtesy of Kenneth Wilson

Wilson said that the new model proposed by Reynolds and his co-authors also helps to understand how phase polyphenism and the associated traits of swarming locusts evolved. "What this model suggests to me is a very logical pathway for trying to understand how these traits evolved in sequence," he said. "All that is needed is that they group together to disrupt the foraging strategy of the predator."

Locusts have plagued mankind since biblical times, descending upon crops in great clouds and devouring everything in their path. Recent outbreaks have vexed farmers in Africa and Australia. Reynolds said that his new model may help managers predict and manage locust outbreaks. "Knowing where [a locust population] is in the environment, and its size and distribution, we can use the model to determine if locust numbers are above or below the threshold for banding and swarm formation," he said. Managers could use this knowledge to alter local environments or vegetation to help prevent outbreaks, Reynolds added.

According to Wilson, percolation theory could be used to describe the behavior of other animals, parasites, or the spread of disease by probing the connectivity of clumped populations and the transfer of pathogens between clumps, for example. "I think it has wider-ranging possibilities than just locusts and predators," he said.

Related story:

Locust Navigation [April 2007]