Abstract. 1. The hypothesis proposed by the fractal foraging model, that prey procurement by predators declines as habitat complexity increases, was tested. To evaluate this hypothesis, the effect of two prey species, second-instar larvae of Scirtothrips perseae Nakahara and Heliothrips haemorrhoidalis (Bouché) (both Thysanoptera: Thripidae), and environmental complexity on the functional response of the predatory thrips Franklinothrips orizabensis Johansen (Thysanoptera: Aeolothripidae) was examined.
2. The Koch curve, a well-studied fractal, was used to generate the shapes of experimental arenas to test the effect of environmental complexity on the functional response of F. orizabensis. Complexity was defined in terms of the number of acute vertices in which prey thrips could seek refuge and the length of the perimeter that had to be searched by the predator. Four shapes were tested: a circle (zero acute vertices, circumference = 186.61 mm), equilateral triangle (three acute vertices, perimeter = 240 mm), star (six acute vertices, perimeter = 277.13 mm), and snowflake (18 acute vertices, perimeter = 333.65 mm). All shapes were scaled so that the area of each arena was 2771 mm2.
3. Franklinothrips orizabensis exhibited a Type II functional response for both prey species and all four experimental arenas tested. Significantly fewer S. perseae larvae were killed in the most complex arena (i.e. the snowflake) when prey densities exceeded 16. For H. haemorrhoidalis, significant differences in mortality were observed only when prey densities equalled four. These results demonstrated that the fractal foraging model was supported when certain prey densities coincided with particular levels of environmental complexity.
4. Subtle changes in environmental complexity and different prey species of varying density can affect prey–predator interactions significantly.