Plant water stress, not termite herbivory, causes Namibia’s fairy circles

The fairy circles of Namibia form a remarkable gap pattern in arid grassland along the Namib Desert. The origin of the fairy circles is subject to an ongoing debate. Solving the mystery of the fairy circles (FCs) requires the right timing in fieldwork after rainfall, as the newly appearing grasses complete their life cycle within only a few weeks. Here we followed the rains along the Namib between 2020 and 2022 and assessed the cause of the grass death within FCs at different time intervals after grass-triggering rainfall. To assess whether termite herbivory was the cause, we used grass excavations and observations on the roots and shoots. To test if edaphic differences may explain the grass death in FCs, we undertook infiltration measurements in 10 FC-hotspot regions. Finally, we used continuous soil-moisture measurements from the dry into the rainy seasons to examine how the newly emerging grasses affect the soil-water content in space and time. Generally, in study plots that received grass-triggering rainfall most recently, the roots of the dead grasses in FCs were in 100 % of the cases undamaged, root-shoot ratios were significantly greater, and the roots were as long or even longer as those of the surrounding matrix grasses outside of the FCs. This indicates that drought stress caused grasses in the FCs to invest resources into roots to reach the percolating water. The results also show that the cause of the grass death in fairy circles was not induced by termite herbivory. Also, we found no systematic differences in the rate of water infiltration between FCs and the matrix, hence the plant wilting cannot result from quicker percolation within FCs. However, the soil-moisture measurements indicate that the matrix grasses strongly depleted the upper soil water of the FCs after rainfall, which explains why most grasses cannot establish and quickly die in the FCs. The research shows that grass death in fairy circles occurs immediately after rainfall due to plant water stress but not due to termite activity. Our results conform with previous fieldwork, pattern analysis and theoretical modeling, suggesting that Namibia’s fairy circles are a self-organized vegetation phenomenon induced by ecohydrological feedbacks.