Scientists crack the code for why locusts swarm

Times of Oman

02-03-2025

Berlin: After months of building, the biggest locust swarm recorded in 70 years swept across 10 countries in East Africa in spring 2020.
The damage to crops was estimated at $8.5 billion (€8.1 billion) in a region where 23 million people face severe food insecurity.
During these invasions, desert locusts (Schistocerca gregaria) eat their own weight in food every day. The biblical-scale plague ate through 160,000,000 kilograms of food a day — enough to feed 800,000 people for a year.
Scientists have been trying to understand how individual locusts gather in swarms for decades. Knowing their behaviour would help with predicting and managing outbreaks.
A new model, published today in the journal Science, casts light on the hive mind of locusts. The study describes how individual locusts transition from behaving as solitary animals to giant swarms with collective motion.
"Our work provides a new perspective for considering collective motion in animals, and robotics too," lead author Iain Couzin, a neurobiologist at Centre of the Advanced Study of Collective Behaviour, Konstanz, Germany.
"One application is a new class of predictive models of how and where swarms move. Future research on this could impact the livelihoods of 1 in 10 people on the planet," Couzin told DW.
A new model of swarming, using insect VR
Locusts swarms have threatened food security for millennia and have played their part in history — locusts were one of the 10 plagues brought upon Egypt as retold in the Book of Exodus.
For decades scientists have been trying to understand how individual locusts move en masse.
In 2006, Couzin developed a model explaining how locusts would march together in a line when they swarm.
"This model came from particle physics and suggested that individuals bump into each other randomly, then flow together all in the same direction if there is a high density of individuals," said Couzin.
Study author Sercan Sayin began probing this model in locusts using a virtual reality (VR) stage set for locusts. Sayin had the insects walk on a ball surrounded by panoramic views on screens. These landscapes reconstructed the world in 3D to make the locusts think they were in a swarm, Sayin said.
But he couldn't replicate the 2006 findings that animal density was responsible for locusts forming swarms.
Vision cues swarming behaviours
Field experiments in Kenya during the huge 2020 swarm showed certain visual cues caused locusts to behave with collective movements when swarming.
"Previously we'd thought that bumping into each other caused swarms, but our experiments showed that it's vision that's important," said Couzin.
"We found instead that [swarm behaviours] are triggered by the type of sensory information around them, not how many locusts they're surrounded by."
Jan Ache, a neurobiologist at the University of Wuerzburg, Germany, who was not involved in the study, said the research expands a mathematical model of swarms which acknowledges the individuality of locusts.
"In order for locusts to have collective motion, they need very basic forms of cognitive processing — where insects integrate their own position relative to the position of those around them, then actively follow other locusts," he said.
This occurs in individual locusts, but when they come together in crowds it creates the emergent effect of a swarm.
How the brain makes decisions
Ache said locusts are fascinating to study because they exist in two different states: solitary or swarming. Normally avoidant, the insects switch into marching bands after several hours of crowding.
"When they change from one type to the other, the brain is in two different states. In each state, the same neurons drive very different behaviors — like being attracted to or repelled by other locusts," Ache said.
Ultimately, the findings are about decisions-making in neuronal systems, Couzin said.
"At the basic level, there's competition between groups of neurons in the brain. The brain must come to a consensus and make a decision about movement," Couzin said.
In other words, when there's a conflict in the brain, neuronal pathways compete until a decision is made when one pathway "wins" over the other.
In their experiments, the visual cues of other locusts in front acted as a target causing the navigation systems to pull the organism in the same direction.
"This is very similar to opinion dynamics in humans, where people adopt similar opinions to others and dismiss other opinions," said Couzin.
Predicting swarms and crowds?
Couzin said the new model has important implications for predicting swarms in the real world.
"If we were able to create a model predicting how swarms move, we were using the wrong model before. The implication is new ways to predict how and where swarms move based on a biological understanding of collective motion," said Couzin.
It could also help to understand how fish move in schools; birds move in flocks and potentially how mammals move in herds. Couzin is also applying their research in robots, creating collective motion in autonomous vehicles.
Couzin said their findings are worth considering in human crowds too, perhaps to help prevent crowd crushes, but "it's too early to make any claims as those experiments haven't been done."