Evolutionary modeling of larval dispersal in blowflies using non-uniform cellular automata

Abstract

When the food supply flnishes, or when the larvae of blowflies complete their development and migrate prior to the total removal of the larval substrate, they disperse to find adequate places for pupation, a process known as post-feeding larval dispersal. Based on experimental data of the Initial and final configuration of the dispersion, the reproduction of such spatio-temporal behavior is achieved here by means of the evolutionary search for cellular automata with a distinct transition rule associated with each cell, also known as a nonuniform cellular automata, and with two states per cell in the lattice. Two-dimensional regular lattices and multivalued states will be considered and a practical question is the necessity of discovering a proper set of transition rules. Given that the number of rules is related to the number of cells in the lattice, the search space is very large and an evolution strategy is then considered to optimize the parameters of the transition rules, with two transition rules per cell. As the parameters to be optimized admit a physical interpretation, the obtained computational model can be analyzed to raise some hypothetical explanation of the observed spatiotemporal behavior. © 2006 IEEE.