ABSTRACT

It is only with the advent of high-speed, large-storage digital computers that
simulations of some complexity have become possible. The development of the present
simulation, which differs in many important respects from other genetic simulations,
and its application to a few specific problems is described. It constitutes a broad,
flexible system for the study of evolutionary problems related to a closed, small
population. Of particular importance are the many similarities to natural populations
such as a time-varying phenotype, a dynamic interaction between the population and
its environment, and mating rules, mortality rules, and selection rules of wide
variety. The definition of the phenotype on the basis of chemical concentrations,
whose levels are under genetic control via enzyme action, is the central feature of
the model.

The relation between the number of crossover positions in the chromosome pair to the
rate of evolution of the population is explored. The importance of the dynamic
relationship between a population under selection pressure and its environment
is underlined. An important result is achieved in the final set of experiments in
which the selection of crossover probabilities is demonstrated. In this effort it
is necessary to apply strong selection pressure to combat the disruptive stochastic
effects in a small population. This pressure is manifested both in more powerful
mortality procedures and in an increased number of allowable offspring per mating.

This system has been developed at least two viewpoints in mind. The first is the desire
to construct a research tool of use to biologists, especially geneticists. The second,
which is dependent on the first, is to illustrate its usefulness by exploring a few
problems of interest to geneticists. With the number of issues discussed and the
importance of the final successful result, it is hoped that the potential of the 
system as a useful research tool is indicated.