Thousands of small, armored crustaceans known as woodlice are marching in tight, continuous loops beneath streetlights at night, unable to break free from circles of their own making. Researchers at The Hebrew University of Jerusalem documented these so-called “death spirals” in the isopod species Armadillo sordidus over a four-year observation period stretching from 2021 to 2025. Image analysis of a single aggregation estimated roughly 5,500 individuals locked in circular movement near one active lamp, and the team identified white light geometry as the driving force behind the behavior.
How white light traps Armadillo sordidus into circular mills
The phenomenon resembles “ant mills,” a well-known pattern in army ants that lose pheromone-trail orientation and march in fatal loops. But this is the first documented case of a similar collective spiral in land-dwelling isopods triggered by artificial light. The study, which appears in the open-access journal Ecology and Evolution, recorded aggregations ranging from thousands to tens of thousands of Armadillo sordidus forming these mills near active streetlights during nighttime hours. The animals are nocturnal and typically avoid light, a trait called negative phototropism that has been studied in isopods since early 20th-century experiments on low-intensity illumination and phototropic reactions.
What makes streetlights different is the geometry of the light source itself. The research team tested whether ultraviolet light or magnetic fields could reproduce the circular behavior and found that both produced negative results. White light from the overhead lamp, by contrast, acted as a fixed spatial reference that appeared to continually redirect the isopods’ forward movement into a curve. Each animal follows simple local rules, turning slightly relative to the light while maintaining contact with neighbors, and the result scales into a massive synchronized spiral.
Work on collective motion in living systems has shown that large-scale circular patterns can emerge from exactly these kinds of minimal individual behaviors without any complex group coordination. In that context, the Armadillo sordidus mills are a textbook example of emergent behavior: no single isopod intends to walk in circles, yet the combination of light-based orientation and crowding produces a coherent rotating structure.
The Hebrew University team described white light geometry as the “smoking gun” for the circular movement. The light does not attract the isopods in a straight line the way a moth flies toward a flame. Instead, it creates a spatial anchor that keeps resetting their directional cues, locking them into a loop they cannot escape on their own. Once a dense ring has formed, the animals at the periphery are hemmed in by conspecifics, and the feedback between crowding and orientation appears to stabilize the mill for hours at a time.
Field data and the 5,500-individual aggregation
The strongest single data point from the study comes from image analysis of one aggregation, which estimated approximately 5,500 individuals in a single mill. That figure, drawn from the study’s supplementary dataset, gives a sense of scale that goes well beyond a laboratory curiosity. Across the 2021 to 2025 observation window, the researchers recorded repeated instances of these formations at field sites with active streetlights, suggesting that the behavior recurs under similar lighting conditions rather than arising from a one-off disturbance.
To arrive at the 5,500-individual estimate, the team used high-resolution photographs and video frames to segment the moving ring into countable units, correcting for overlap and motion blur. While the exact error margin is not specified, the order of magnitude is clear: this is a crowd of thousands of animals, not hundreds. In some sequences, multiple concentric rings appeared to rotate in the same direction, hinting at even larger total counts when outer, less-dense bands are included.
The study does not report direct mortality counts inside the mills. The primary datasets, including raw CSV files and video recordings archived in the journal’s online materials, contain individual counts and movement data but no systematic measurements of predation rates or dehydration deaths within the spirals. The term “death spiral” comes from the broader biological analogy to ant mills, where trapped individuals often die of exhaustion. Whether the same lethal outcome consistently applies to Armadillo sordidus remains an open question in the data.
Anecdotal field notes in the study mention sluggish movement and scattered carcasses near some former mill locations the following morning, but these observations are not quantified. Without controlled comparisons between mill and non-mill sites, it is impossible to say how much additional mortality the spirals cause beyond normal background levels. Predators such as ground beetles or birds might also exploit the dense aggregations, but again, no direct measurements are available.
Separate research on LED streetlight effects has shown that artificial light at night measurably alters the functional composition of ground-dwelling invertebrate communities. That broader body of evidence supports the idea that the isopod mills are not an isolated oddity but part of a wider pattern in which urban lighting reshapes nocturnal invertebrate behavior at the community level. In that context, even modest additional mortality or displacement of Armadillo sordidus could ripple through local food webs.
Unanswered questions about light intensity and species range
One of the sharpest gaps in the current evidence involves the specific properties of the streetlights that trigger the mills. The published study does not report the color temperature, spectral output, or lux intensity of the lamps at each field site. A plausible hypothesis is that mills form only when streetlight color temperature exceeds a threshold and intensity stays within a narrow band strong enough to override the animals’ natural light avoidance but weak enough to still permit forward locomotion. That threshold could be tested through controlled lamp swaps at matched field sites, but no such experiment has been reported.
Laboratory trials under programmable LEDs would allow researchers to vary brightness and spectrum independently, tracking when straight-line escape behavior gives way to curved paths and, finally, to stable circular mills. Such experiments could also test whether polarized light patterns from different lamp housings influence orientation, or whether the crucial factor is simply a bright, fixed point overhead.
Geographic scope is another limitation. All documented observations come from Armadillo sordidus populations, and no broader surveys or institutional biodiversity databases have confirmed whether the behavior occurs in other isopod species or in regions outside this species’ known range. Without multi-site, multi-species data, it is difficult to estimate how widespread the phenomenon might be as urban lighting continues to expand.
Targeted surveys at streetlit sites in other cities and habitats could quickly reveal whether related woodlice show comparable mills or whether Armadillo sordidus has an unusually strong sensitivity to overhead white light. If the behavior proves common, it would strengthen the case for considering isopods in lighting impact assessments, alongside more frequently studied groups such as moths and bats. If it remains restricted to a narrow set of conditions, the spirals would still offer a striking natural experiment in how simple sensory rules interact with human infrastructure.
Implications for urban lighting and future research
For now, the Armadillo sordidus spirals underscore how little is known about the fine-scale behavioral effects of streetlighting on small ground fauna. A standard engineering decision-installing bright, broad-spectrum lamps for human safety and visibility-has unintentionally created a setting where thousands of invertebrates can become trapped in endless circles. Whether or not those circles are consistently lethal, they represent a profound alteration of normal nocturnal activity patterns.
Adjusting lamp spectra toward warmer tones, dimming lights during low-traffic hours, or adding shielding to reduce downward glare are all existing tools in urban lighting design. The new findings suggest that isopods and other ground-dwelling invertebrates should be included when evaluating which combinations of these tools best balance human needs with ecological impacts. Future work that links detailed behavioral observations to specific lighting technologies will be key to turning the eerie image of woodlice death spirals into actionable guidance for city planners.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
