Christopher
Cosner
Department of Computer Science
University of Miami
gcc@cs.miami.edu
This talk will describe an approach to modeling a predator-prey
system which involves multiple spatial and temporal scales. The
models describe a natural system where the predators are ladybird
beetles and the prey are aphids. The underlying environment consists
of a collection of patches of plants which may support colonies
of aphids. The beetles disperse through the environment quite
rapidly by searching the patches for aphid colonies, but reproduce
only once per season. The aphids colonize some of the patches
and go through several generations per season, but disperse slowly
within each patch. Thus, the predators are assumed to interact
with the environment on a large spatial scale and to experience
it as a collection of patches, to disperse on a very fast time
scale, and to reproduce on a slow time scale. The prey are assumed
to interact with the environment on a small spatial scale and
to experience each patch as a continuum, and to disperse and reproduce
on moderately fast time scales. Structurally the models are a
hybrid of patch models (for the predators) and reaction-diffusion
models (for the prey). The assumptions about time scales are used
to simplify the models. The models can be used to study the effect
of patch size on population dynamics. In some parameter ranges
the models predict that smaller habitat patches will sustain higher
prey densities than larger patches. This phenomenon has been observed
in a natural system consisting of aphids and ladybird beetles.
It is an emergent propery of the system in the sense that models
based on the prey alone typically predict higher population densities
on larger patches. Other issues such as comparisons of different
foraging strategies which might be used by the predators can also
be addressed from the same general modeling viewpoint.
Back
to Workshop Schedule
Back
to Biocomplexity Workshop
1999-2000
Reactive Flow and Transport Phenomena