Ecologists, evolutionary biologists and engineers all share an interest in
the causal relationships between individual organisms' social
interactions, group characteristics, and the distribution of populations
among groups. Much of the interest, and also the difficulty, of studying
biological grouping stems from how hard it is to simultaneously quantify
dynamics at the individual-, group- and population- levels. Combining
microscopic individual- and group-level observations with population-level
simulation tools is a natural strategy. In this talk, we will use detailed
analysis of laboratory-observed individual-level fish behavior in distinct
types of groups to hypothesize novel interaction rules for schooling
simulations. We will then examine some population-level implications of
hypothesized behavioral types, using large-scale simulations, mathematical
analysis, and "equation-free modeling" approaches.
(*Co-author, Jeff Moehlis, University of
California, Santa Barbara)