SUMMARY
Intensive shrimp farming is a lucrative and highly risky business. Before entering this industry, most farmers spend time observing the operation of pilot farms. This stage is important to master essential techniques and judge the profitability and risk associated with shrimp farming. Learning is a complex process that leads to misconceptions about the nature of short-term and long-term risks. This paper uses computer simulation to illuminate the dynamic nature of the learning processes, land conversion, shrimp production and environmental contamination. The model is based on conditions of the Dai Hoa Loc Commune in the Mekong Delta of Vietnam. Initial simulations match statistical data by revealing the high risk: high initial profits from the pilot farms followed by conversion from rice land to shrimp farms. When rapid conversion occurs, the region is vulnerable to excessive accumulation of nutrients, a decline in shrimp yields and financial failure. In contrast, low stock densities deliver a lower profit which is insufficient to stimulate mass land conversion. The paper concludes with testing recovery strategies for farmers who have suffered the longer term impacts of high stocking density. Results show that yield recovery is possible by improving the channel and imposing regulatory control over stocking density.