The aim of this study was to provide a framework, for describing and understanding the geometry of movement of insects foraging within complex plant canopies where the insect is exposed to varying stimuli. We used the apple maggot fly, Rhagolttis pomontUa (Walsh) (Diptera: Tephritidae), foraging in apple trees devoid of fruit as our model system. The framework provides the null hypothesis required for inferring the influence of external stimuli, such as fruit color and odor, on the paths of

more »

... ging flies. We mapped trees into cells, released preconditioned flies in caged trees, and recorded their behavior and location. Flies moved mainly to the nearest neighbor cells, but displacement within a wide range of distances was observed. The model closest to observations is a random walk with a position-dependent bias in the vertical component of movement. Four other models were built, spanning a range of simplifications in the rules determining the vertical component of movement We used the concept of avoidance of self-crossing in a searching path for defining efficiency of movement Flies were quite efficient at visiting almost as many different cells as possible. Comparisons of assumptions and predictions of the five models revealed that this efficiency is due to the small number of steps, the location of the starting cell, and a strong tendency to move upward in its vicinity. We discuss the selection pressures on movement rules: pressure from predators may explain the short hops, while the sensory ecology of fruit finding and the avoidance of sites already visited by other flies or by the same fly may explain the position-dependent upward bias. Strong similarities between the rules for die vertical component of movement of one simplified model and the observations lead us to believe that canopy architecture influences insect movement not only by defining the set of locations that the insect can visit using predefined rules for movement but also by defining die rules of movement Kty words: apple fruit fly, canopy, foraging behavior, fruit fly, movement, plant architecture, RhagoUHs pomontUa, searching efficiency, search theory. (Behav Ecol 8:37-45 (1997)] I nsects typically search for food, hosts, or sexual partners in geometrically complex environments. Examples include the flight of insects within structurally variable pheromone plumes (Baker, 1989; David, 1986; Elkinton et aL, 1987; Murlis and Jones, 1981), the movement of foraging coccinellids and parasitoids on plants (Ayal, 1987; Brodeur and McNeil, 1991; Casas, 1991; Frazer and McGregor, 1994; Hoffmann, 1990), and die movement of darkling beetles in semiarid grasslands (Johnson et al ., 1992) . Nevertheless, modeling insect movement has mainly been restricted to continuous homogeneous environments (Kareiva, 1990). The reasons for assuming a homogeneous environment when modeling and analyzing insect movement have their origins in two decisive advantages. First, a body of analytical solutions can be relied on, mainly centered around the simple diffusion equation. The issues addressed by Okubo (1980) .provide a good representative example of this approach. Second, insect movement can also be studied without paying attention to the fine structure of the environment This practice rests on the lesuicuve assumption of deterministic homogeneous geometry of the texture of the medium at die microscale. Its corollary is that die large-scale results of diffusionary processes are virtually independent of die fine texture of die medium (Skellam, 1979:67-68). Incorporation of even mild disorder in die environment can lead to unexpected results about die dispersion of moving particles, such as a nonlinear increase of the mean-square displacement with time, which constitutes a severe violation of die basic laws of die diffusion equation. Only a few ecological studies do justice to heterogeneity in J. Cuas is now at IRBI-CNRS UPRES A 6035. Avenue