Scientists seek locust plague predictors

Interactions leading to cannibalistic behavior may be key to pests’ control
Writer: Steve Byrns, 325-653-4576,
Contact: Dr. Gregory Sword, 979-862-1702,

COLLEGE STATION – Since ancient times, huge, angry clouds of hungry locusts laying waste to the countryside have been synonymous with human despair and suffering, but now a team of scientists may have the key to stopping the hordes before they start.

Some locusts such as these African desert locust (Schistocerca gregaria) juveniles, change color as well as their behavior when crowded. The green locust is in the solitarious phase; the yellow and black is in the gregarious phase. (photo by Dr. Greg Sword)

Dr. Greg Sword, Charles R. Parencia chair in cotton entomology at Texas A&M University, is part of the team who recently published the paper “Cannibalism can drive the evolution of behavioural phase polyphenism in locusts” in the scientific journal “Ecology Letters.”

The paper interprets their work using artificial locust simulations to explain the insects’ destructive nature. The team hopes to ultimately develop computer-driven predictive locust movement models to predict and stop locusts from swarming, Sword said.

Sword is quick to point out that the locusts the team dealt with are not the cicadas most Texans hear singing on a hot summer afternoon nor are they any of the several grasshopper species that have plagued Texans this summer. The locusts Sword refers to are members of the genus Schistocerca and others that inflict mass devastation over large parts of Africa, Australia, Asia, and Central and South America causing untold destruction and human suffering.

The international scientific team studying the insects includes Vishwesha Guttal, formerly at Princeton University, now at the Indian Institute of Science, Bengaluru, India; Pawel Romanczuk, Max Planck Institute for Physics of Complex Systems in Dresden, Germany, Stephen J. Simpson, University of Sydney, Australia; Iain D. Couzin, Princeton University and Sword.

Australian plague locust juveniles (Chortoicetes terminifera) during an outbreak in central New South Wales, Australia. (photo by Dr. Greg Sword)

“If we can predict where locust swarms will be at a given point in time, we can minimize the amount of pesticide and manpower needed to control them, and possibly minimize the damage they cause,” Sword said. “Targeting migratory bands of juveniles on the ground before they become flying swarms that can travel huge distances is a good start, but predicting their movement has been notoriously difficult.”

Sword said having good knowledge of how locusts interact with each other will ultimately allow the team to model the swarming and movement process using models based on real locust behavior.

He said locusts are grasshopper species that harbor a most sinister trait; they can change from placid individuals to ferocious cannibalistic eating machines when crowded. This “Jekyl and Hyde” transformation can occur in as little as four hours. The technical term for the change is behavioral phase polyphenism and it occurs in all locusts.

“Lone locusts are very sedentary and actually avoid each other,” Sword said. “This is called the solitarious phase. However, when they become crowded, as happens under outbreak conditions, they can undergo a rapid behavioral shift and suddenly become active and move toward each other. This is called the gregarious phase. This dramatic change is the defining characteristic of locust biology. It’s what makes a locust a locust and not just another grasshopper.”

“Most people might think the insects change behavior to swarm much like bees absconding from an established hive to create a new one, but locusts are actually at the opposite extreme. They are desperately trying to escape fellow locusts who are trying their best to make a meal of them and anything else that gets in their way.

“Our work shows that both the low and high density strategies the insects use evolve to minimize the risk of contacting other locusts and being eaten. When there’s just a few locusts, avoidance always works, but as populations increase, so do collisions and the risk of being munched on,” he said.

“This is when the switch to fleeing locusts coming from behind and moving toward those moving away becomes favored to avoid cannibalism. The result are the dreaded huge moving swarms these insects are so well known for,” Sword said.


“They sure aren’t migrating in the typical sense of the word. They are trying to escape each other. Of course, they are hungry and must eat too, resulting in mobile swaths of mass destruction.”

Sword said the “safety in numbers” strategy to avoid predation has also been studied and while that may occur, the presence of predators is not necessary for the swarm to develop.

“Our study used some pretty complex and high-powered evolutionary simulation computer modeling to test how locust behaviors could evolve at different population densities under the risk of cannibalism,” Sword said.

“The models described very closely the real locust behaviors observed in nature. Thus, not only can cannibalism trigger the mass migration in crowded locusts, we also learned that cannibalism is likely what caused their behavioral changes to evolve in the first place. These changes center around seeing one another in close proximity and actually being bumped by their neighbors. To a locust, both are potentially precursors to being eaten.”

Sword said another important part of their study showed it’s much harder to stop locust outbreaks once they enter the destructive gregarious phase than it is to prevent them from reaching that point.

“This means if we wait until outbreaks develop as we usually do, we then have to reduce locust numbers way below the number needed to prevent the locusts from swarming in the first place,” he said.

“It’s much like fire prevention; better and much less costly in the long run to keep it from starting than to control it after it has escaped. Controlling locust outbreaks in Africa, for example, can cost hundreds of millions of dollars in direct assistance and famine relief, so the potential savings that preventative measures based on predictive movement models has the potential to deliver in the future could be substantial.”

To view the paper, go to:


Print Friendly, PDF & Email