The Cook Book
Recipe for the week of January 29 - February 4
CAM8 as an Ecomachine: 1 Predator, 2 Prey
- This week's soup depicts one of our first CAM8-based ecology modeling
experiments, as reported at the Santa Fe Institute
workshop last month. A 512 by 512 array of 'patches' is populated by three
species: the Good prey (benign), the Bad prey (pest), and the Ugly predator. These
creatures move around, give birth to additional members of the same species on the
same patch, and die. The birth, death and transport rates for a creature can all
depend on the demographics of its patch. Without going into details, this framework
permits the empirical investigation of a rich class of nonlinear competitive random
CA dynamics. The performance of our simulation on CAM8 is about 400 times that of a
Power Mac program, providing real-time interactive visualization of systems large
enough to exhibit complex spatial structure.
- The color scheme for our implementation is unforgettable. The Good guys are
green, every ecologist's favorite color. The pests are blue meanies, and the
bloodthirsty predators are red. Our three species inhabit each
patch with one of four densities: zero, low, medium or high, with the intensity of
their colors adjusted accordingly, and with mixed populations being represented in
standard RGB coordinates. For instance, bright yellow corresponds to a patch with a high
density of predators and benign prey but no pests. Our soup shows the state of the
ecology a few hundred time steps after a small density of predators has
been introduced into an equilibrium of good and bad prey in which the pests
predominate. At this early stage, wave fronts of the predators reminiscent
of simpler 'forest fire' dynamics are clearly visible.
- One interesting story connected with this experiment may be viewed as a
cartoon of biocontrol. For certain parameter values discovered by
Eric Klopfer, it turns out that the pests predominate in the absence of
predators, the predators are unable to eliminate the pests in the absence
of good prey, but the pests can be eliminated without seriously impacting the
good prey if all three species are present. We plan to report on further
developments with this model, a corresponding two-predator, one-prey
system, and other Ecomachine simulations in future recipes.
|
