Fractal index captures the role of vegetation clumping in locust swarming

1. This paper evaluates box-counting dimension as a quantitative clumping index for discontinuous plant cover, and applies it to studies of both small- and large-scale ecological processes in desert locust (Schistocerca gregaria) swarming. 2. The quadrat under study is tiled with squares of increasing size, and the relationship between the number of occupied squares and square size gives the box-counting dimension. This index is high for random distributions and decreases when the vegetation is clumped; it is also positively linked to the abundance of vegetation. 3. At the scale of individual locusts, this index captures the space-filling properties of vegetation that promote change to the gregarious phase and the onset of swarming. At the very large scale, box-counting dimension measured from satellite imagery reflects topography, which influences the concentration and migration of locust swarms. 4. Fractal geometry considers scale of measurement in quantifying pattern and provides a framework to interpret ecological phenomena across spatial scales. Demonstration of a statistically significant relationship between the box-counting exponent and locust behaviour shows how this approach can be applied to ecological studies.