SDF - A phenomenological model for scale-dependent foraging
Imagine insects arriving in a unit size arena and deciding where to alight. Initially, take n circular samples (i.e. snapshot views) of the arena, each sample being randomly placed and of diameter s (where s < 1). Next, “record” the number of plants sensed in each sample and zoom-in on the sample with the highest (or lowest) plant density. Within the reduced area take n finer-scale samples of diameter s2. Keep selecting the best of n subsamples and reducing the sample diameter by a factor of s for five or more iterations (or until all samples contain either 0 or 1 host plant) then select the plant nearest to current location. If n = 1, the model equates to picking a random location and moving to the nearest plant, hence isolated and edge plants are favoured because they have large “catchment areas” (Fig 1A & 2A). Increasing n creates a more pronounced and deterministic response (Fig 2B-D). Parameter s determines the relative scale at which patch density is assessed, and this may be altered as the insect zooms-in (e.g. Fig 2D), if s is initially very small all plants have a similar probability of being chosen, while large values of s favour plants in large-scale clumps.
Results
from a typical run are shown below.
Dots represent plants and circle
size represents the relative number of insects selecting each plant.
Download User-Friendly Simulation: Click here to “Open” (i.e. Run) or “Save” the Visual
Basic executable file
Instructions
When getting started use the
default parameters with Running commentary = ON
With increasing familiarity,
turn off the Green and Red display and the Running commentary and increase the
number of foragers