Virtual reality reveals the true way of thinking of locusts in large swarms

As millions of desert locusts move together in the North African landscape, consuming everything green, they seem to share a thought. For decades, scientists believe these spectacular groups have followed simple rules based on consistency with their neighbors. New research using virtual reality has just put this understanding on its head, revealing that locusts are complex decision makers, not particles that follow fixed rules.

The groundbreaking study, published in Science on February 28, challenges long-term theories on how locusts coordinate their movements, on predicting and managing the effects of devastating outbreaks that threaten food security across multiple continents.

“It is time to transcend the concepts of locusts and other organisms because particles acting according to fixed spatial and temporal rules and treat organisms as probabilistic decision makers that dynamically respond to their sensory environment.”

For more than two decades, scientists have relied on models borrowed from physics (original propelled particles (SPP) models) to explain how locusts achieve their significant synchronization. This approach shows that as the density increases, the locust will automatically align its direction of movement with nearby neighbors, triggering a sudden shift from chaotic movement to coordinated march.

The research team led by Sercan Sayin solved a significant technical challenge that limited previous research. Rather than trying to infer individual locust behavior from group observations, they created an immersive virtual environment where individual locusts can move freely while being surrounded by computer-generated locusts.

As Buhl and Simpson described, this “technology tour” allows researchers to precisely control what each test’s lens sees while measuring how it responds, which is not possible in a natural cluster where everyone affects others at the same time.

What they found fundamentally changed our understanding of these ancient agricultural pests. When the two groups were moved in the same direction, the test locust did not continue to align the motion as predicted by all previous models. Instead, they determine the decisive turn of one or another group, viewing other locusts as goals that pursue rather than being consistent with their neighbors.

Researchers have not found evidence of density thresholds that trigger coordinated motion, another core prediction of traditional models. Furthermore, locusts do not respond to wide-field optical flows as proposed by some alternative theories.

These findings point to what neuroscientists call the “ring attractor model”, where the locust’s brain constantly updates its title based on changing visual inputs and internal neurodynamics, similar to the navigation system found in other insects.

The Desert Locust (Schistocerca Gregaria) is perhaps the most dramatic example of nature’s behavioral plasticity. When raised in isolation, these insects avoid each other and try to hide in predators. However, only hours of crowding triggered a significant shift that led them to search for other locusts and form dense agglomerations that eventually turned into the parade bands that farmers fear throughout the history of the recording.

It is not just academic curiosity to understand exactly how these groups work. Locust outbreaks remain one of the most lasting threats to agriculture, capable of consuming crops that feed millions of people. The 2019-2022 locust crisis in East Africa and South Asia affected more than 23 million people, causing control efforts and crop losses to hundreds of millions.

The next research challenge will be to test whether this new cognitive model can successfully predict the movement patterns of actual locusts across larger landscapes and real-world environments. The researchers will also study whether the model explains the specific localization of locusts in the frequency band, which previous studies have shown are the following unique patterns.

As climate change may change the frequency and intensity of locust outbreaks, this deeper understanding of population dynamics can help authorities develop more effective early intervention strategies that save millions of food resources and control costs.

This study represents a fusion of behavioral analysis, neuroscience and virtual reality technologies that ultimately allow scientists to gaze at the cognitive processes that drive one of nature’s most fascinating and economically important collective behaviors.