Mysterious 'fairy circles' in African desert: Here's a new theory wrote:The bizarre circular patches of bare land called "fairy circles" in the grasslands of Africa's Namib Desert have defied explanation, with hypotheses ranging from ants to termites to grass-killing gas that seeps out of the soil. But the patches may be the natural result of the subsurface competition for resources among plants, new research suggests.

Grasslands in the Namib Desert start off homogenous, but sparse rainfall and nutrient-poor soil spark intense competition between the grasses, according to the new theory. Strong grasses sap all of the water and nutrients from the soil, causing their weaker neighbors to die and a barren gap to form in the landscape.

The vegetation gap expands as the competition ensues, and the grass-free zone becomes a reservoir for nutrients and water. With the additional resources, larger grass species are then able to take root at the periphery of the gap, and a stable fairy circle develops. [See Photos of Mysterious Fairy Circles of the Namib Desert]

"It's a really good theory because it accounts for all the characteristics of fairy circles," including the presence of tall grass species, Florida State University biologist Walter Tschinkel, who was not involved in the study, told LiveScience. "No other proposed cause for fairy circles has ever done that."

A lingering mystery
Fairy circles have been a mystery to scientists for decades. Last year, Tschinkel discovered that small fairy circles last for an average of 24 years, whereas larger circles can stick around for up to 75 years. However, his research didn't determine why the circles form in the first place, or why they disappear.

Earlier this year, University of Hamburg biologist Norbert Juergens claimed to have found evidence for a termite theory of fairy circles. Essentially, he discovered colonies of the sand termite, Psammotermes allocerus, were nearly always found in the centers of fairy circles, where he also found increased soil moisture. He reasoned that the termites feed on the grasses' roots, killing the plants, which usually use up the soil's water, and then slurp up the water in the resulting circular patches to survive during the dry season.

But Tschinkel is critical of the work, stressing that Juergens confused correlation with causation.

Michael Cramer, a biologist at the University of Cape Town in South Africa and lead researcher of the current study, which was published recently in the journal PLOS ONE, also thinks the termite theory falls short.

"I think the major hurdle that explanations have to overcome is explaining the regular spacing of the circles, their approximate circularity and their size," Cramer told LiveScience. "There's no real reason why termites would produce such large circles that are so evenly spaced."

Scientists have also previously proposed that fairy circles are an example of a "self-organizing vegetation pattern," which arises from plant interactions. In 2008, researchers developed a mathematical model showing the vegetation patterning of fairy circles could depend on water availability.

A fierce competition
To test this theory, Cramer and his colleague Nichole Barger from the University of Colorado at Boulder first measured the size, density and landscape occupancy of fairy circle sites across Namibia, using both Google Earth and ground surveys. They then collected soil samples at various depths from inside and outside the circles, and analyzed them for water and nutrient content. Finally, they plugged the information, along with climate data such as seasonal precipitation and temperatures, into their computer models. [Images: The 10 Strangest Sights on Google Earth]

"We found that the size of the circle, the density and degree to which they occupy the landscape are all associated with the amount of resources available," Cramer said. Specifically, fairy circles are smaller if they have more resources, such as soil nitrogen and rainfall.

This makes sense, Cramer explained, because the taller grasses won't need a large reservoir of resources to get started and survive if water and nutrients are already available in the environment. On the other hand, the grasses require a large reservoir to sustain themselves if the soil is poor in water and nutrients.

The researchers also discovered that rainfall strongly determines the distribution of the fairy circles across Namibia, with circles only appearing in areas where there is just the right amount of rain (not too little, but not too much). If there's too much rain, the bountiful resources would "relax" the competition for resources and the circles would close up; but if there's too little rain, the competition would become too severe and the circles would again disappear, Cramer said. Because the circles can occur only in this narrow moisture range, differences in rainfall from year to year may cause them to suddenly disappear and reappear in an area over time. With this information, they found that they could predict the distribution of the fairy circles with 95 percent accuracy.

Additionally, the regular spacing between fairy circles may be the result of inter-circle competition, with grasses from each circle "battling" with other circle grasses for resources, Cramer said.

Experimental tests
Cramer notes that termites may still be involved in fairy circles. "What sets up the circles is the competition between plants," he said. "Termites are a secondary phenomenon, and their role is to serve as a maintenance for the circles by killing off the grasses that spring up in the center of the circles."

Yvette Naudé, a chemist at the University of Pretoria, South Africa, who was not involved in the study, thinks it's refreshing to see a noninsect hypothesis for fairy circles, though she expressed some doubts about its validity.

"It is unclear how peripheral grass resource-competition could induce such abrupt and synchronized plant mortality over an entire patch," Naudé, who has previously studied fairy circles, told LiveScience in an email. (Cramer actually thinks the plant mortality starts off small, and the patch grows as the competition continues.) "The answer to the enigma (of fairy circles) remains elsewhere."

To examine whether the theory is correct, Cramer plans to conduct experimental tests, as his study provides only correlative evidence for the competition theory.

"If fairy circles really do develop from a shortage of water and nutrients, then simply watering and fertilizing the circles should cause them to close up with vegetation," Tschinkel said.