In this research the population dynamics of Paramecium caudatum isolated from an artificial wetlands system and culture in laboratory was modeled by means of Cellular Automatons (CA). The experimental conditions of culture of the isolated Paramecium's were carried out in regular media and Cu+2 media; the last one for toxicity bioassays. The CA was developed with the direct problem approach, which consist of laboratory experiments and then fits a pseudo-random rule to achieve a similar dynamics to the real system. Thus, a bi-dimensional array of 600x600 cells, which could be contained 0 or 1 value; this was taken as a representation of the distribution of the microorganisms according to the presence or absence of 0 to 1 respectively. At the same time, each cell had further information related to substratum consuming and threshold division of the P. caudatum. In addition, the Moore's neighborhood' was utilized in the problem solution. The kinetics' equations for biological growth were applied to transition rules and determined in a direct way. Each pulse or evolutionary step in the computational system was carried out by the hour for all cells in that current state. The comparative result of silicon simulations with the laboratory experiments of this kind of modeling shows clearly the utility of the evolutionary computation in the re-creation of real biological dynamics. The model developed, was simulated by means of a Java Applet with a friendly interface, which allows the adequate interaction with the user.