The stock model, together with the estimates of regression coefficients, was used to predict the dynamics of production for hypothesized alternative management strategies that considered the role of stocking density, pond size, and starting time of aeration. The extreme (minimum and maximum) values of the management variables and the mean values of water quality variables contained in the database were used for prediction.