From the air, parts of the Namib Desert look as if someone has stamped an enormous sheet of polka dots across the landscape. Thousands upon thousands of roughly circular patches of bare ground interrupt the grassland, often with a taller fringe of grass around their edges. They can persist for years, disappear after unusually wet periods and reappear as drought returns.
They are known as fairy circles, although there is nothing supernatural about the two explanations that have dominated decades of research. One camp argues that tiny subterranean engineers — sand termites — kill vegetation and maintain the bare patches. Another says the pattern emerges spontaneously because grasses competing for desperately scarce water organize themselves across the landscape.
The disagreement has become unusually vigorous for a desert ecology problem. Papers have directly challenged one another's measurements and interpretations, researchers have returned to the same landscapes after rainfall, and new studies continue to support different parts of the competing stories.
As of 2026, the most accurate answer to the deceptively simple question “What makes a fairy circle?” is not that science has failed. It is that scientists have narrowed the mystery to mechanisms precise enough to test — and they still disagree about which one is doing the crucial work.
A landscape of remarkably orderly holes
Namibian fairy circles occur through an arid belt along the eastern margin of the Namib, extending through parts of southwestern Africa. A typical circle consists of a bare central patch surrounded by perennial grasses, often species of Stipagrostis. Diameters commonly fall within a few meters, although circles can be considerably larger.
Their most striking property is not merely their shape but their spacing. In many landscapes the circles are “overdispersed”: neighboring gaps tend to keep a characteristic distance from one another rather than being scattered randomly. Viewed across a large area, the result resembles an enormous natural honeycomb without straight edges.
That regularity is a major clue. Whatever mechanism makes the circles must somehow operate over distances larger than an individual grass plant.
Older explanations ranged from toxic plants and soil chemistry to radioactive ground, gas seepage and the activity of various animals. Most have failed to explain the full combination of bare centers, grass fringes, geographical distribution, long persistence and striking spatial order.
Two explanations survived as serious competitors.
The termite hypothesis: invisible ecosystem engineers
The animal hypothesis focuses particularly on the sand termite Psammotermes allocerus. These termites live underground and can feed on plant roots. Ecologist Norbert Jürgens and collaborators have argued that their colonies deliberately or indirectly create the bare patches by removing vegetation.
The proposed mechanism is elegant. If termites kill grasses in a central area, the bare soil loses less water through plant transpiration. Rain can infiltrate the sandy ground and remain available below the dry surface. The circle becomes an underground moisture reservoir that helps the termite colony survive long periods without rain.
The ring of grasses around the perimeter can then provide food. Competition between neighboring termite colonies could contribute to the regular spacing of circles across the landscape.
Supporters of this interpretation have reported strong associations between fairy circles and Psammotermes, moisture persisting beneath bare patches and damage to grass roots that they attribute to termite feeding. A 2023 Royal Society study described the insects as ecosystem engineers whose colonies create water-storing bare areas in the hyperarid Namib.
In a direct 2023 response to critics, Jürgens and soil scientist Walter Gröngröft argued that grasses in the bare centers die while subsurface soil is still moist, that horizontal movement of water through the sand is insufficient to explain the deaths and that root herbivory provides the missing cause.
If they are right, the circles are essentially termite-built water infrastructure on a landscape scale.
The plant hypothesis: a pattern that organizes itself
The rival explanation requires no animal architect.
In drylands, plants face a spatial problem. A continuous carpet of grass would demand more water than the environment can reliably provide. Mathematical models of vegetation show that competition and facilitation can cause plants to self-organize into regular spots, stripes, labyrinths or gaps.
Applied to Namibia, the idea is that grasses around a developing bare patch gain access to soil water that would otherwise support plants in the center. Their roots draw water laterally and vertically through the sandy substrate. The successful grasses grow more strongly, while seedlings inside the increasingly depleted topsoil wilt. The gap therefore reinforces the vegetation around it.
No grass needs to “decide” where a circle belongs. Local interactions repeated across thousands of plants can create an ordered pattern at landscape scale, much as simple physical rules can produce ripples in sand or regular convection cells in a heated fluid.
Stephan Getzin and colleagues have produced some of the strongest field evidence for this view. Between 2020 and 2022, they followed rainfall events across multiple Namibian fairy-circle regions and excavated young grasses shortly after germination. They reported that grasses dying inside circles had intact roots rather than the herbivore damage expected if termites had killed them. Soil-moisture measurements indicated that the surrounding grasses rapidly depleted water available to seedlings in the circle's upper soil.
In their interpretation, the seedlings were dying of thirst.
A 2024 follow-up examined drone-mapped plots and rainfall responses. The most regularly ordered circles occurred in deep, homogeneous aeolian sands, where water infiltrates efficiently and plant competition can operate symmetrically. After abundant rain following drought, hundreds of mature circles temporarily filled with grass, vastly outnumbering newly formed gaps. That dynamic response to water strongly supported the authors' self-organization model.
Why can both sides look at water and reach opposite conclusions?
This is the heart of the controversy.
Both camps agree that water is central. The Namib is an intensely water-limited environment, and fairy-circle soils can retain moisture below the surface. What they dispute is why that water distribution exists and what kills the grasses attempting to grow in the center.
Termite proponents argue that insects first remove vegetation. The absence of plants then conserves water, producing the moist reservoir beneath the circle. In that sequence, termites create the gap and water storage follows.
Self-organization proponents reverse the causal logic. Competition among grasses redistributes scarce water and causes seedlings inside the gaps to desiccate. The bare patch and its moisture pattern are emergent consequences of vegetation dynamics rather than insect engineering.
Even soil depth matters. One layer can be relatively dry while deeper sand remains moist. A newly germinated grass with short roots may die because its shallow root zone is dry even if substantial water exists farther below. That distinction has fueled direct exchanges between the research groups over hydraulic conductivity, root length and the timing of measurements after rain.
The dispute is therefore not simply “termites versus water.” Both theories involve water and biological feedback. The real question is which organism drives the feedback that initiates and maintains the circle.
What does the newest evidence say?
A 2025 review of fairy-circle research concluded that the debate remains open but identified weaknesses that future experiments need to address. It noted that termites are not ubiquitous across every documented circle and that their presence alone cannot automatically establish causation. At the same time, ecological self-organization can reproduce many observed patterns but still requires more direct experimental demonstration of the proposed underground water flows and competitive effects.
The authors called for manipulative tests rather than another round of correlations: remove or suppress termites in controlled plots and see whether circles persist; manipulate soil water and nutrients and measure how vegetation responds; track water fluxes and roots directly through the critical weeks after rain.
That is a crucial point. Finding termites beneath a circle is not enough, because termites may prefer a habitat created by the circle. Finding a water pattern predicted by a vegetation model is not enough either, because termite activity could have produced that water pattern. Cause and consequence can imitate each other.
Other organisms add further complexity. A 2025 study sampled soil nematodes along roughly 900 kilometers of Namibian fairy-circle habitat. Bare centers had very low nematode abundance and diversity compared with grassy rings and surrounding vegetation. The result did not solve the origin question, but it showed that once a circle exists it develops a distinctive belowground ecology that may help sustain its conditions.
In other words, the mature circle may be an ecosystem with multiple feedbacks even if one mechanism originally triggered it.
Could termites and plants both be involved?
A combined explanation is tempting. Termites might initiate or reinforce some gaps while plant-water feedbacks organize and stabilize the larger pattern. Ecological systems often have multiple causes, and mechanisms need not be mutually exclusive in every location.
But “both” cannot simply be used as a comfortable escape from testing. Scientists still need to establish which process is necessary, which is sufficient and whether their importance changes with soil, rainfall and geography.
There is also a danger in assuming that every circular vegetation gap on Earth is the same phenomenon. Fairy-circle-like patterns have been reported outside Namibia, most famously in arid Western Australia, and global satellite studies have identified many landscapes with similar-looking regular gaps. Visual resemblance does not guarantee a shared cause. Different ecosystems can converge on similar patterns through different mechanisms.
That is why Namibia remains so valuable. Researchers can return after rare rainfall, measure the first hours and days of seedling growth, excavate roots, map termite activity, monitor moisture at different depths and use drones to watch entire landscapes reorganize.
The circles are not static marks etched into the desert. They expand, contract, revegetate and respond to sequences of wet and dry years. Their apparent perfection is the temporary visible result of a living system under extreme pressure.
Perhaps that is the real reason the mystery has lasted. A fairy circle looks simple: bare sand in the middle, grass around the edge. But beneath that geometry lies a contest among water, roots, insects, soil physics and time. Whether termites or thirsty grasses ultimately win the scientific argument, the Namib has already demonstrated something more surprising — an ecosystem can turn scarcity itself into a pattern visible from the sky.
