Management of natural resources and the optimization of their uses is a complex problem due to the need for an intertemporal solution and because the resource stocks are usually not known with certainty. Cases • such as the stock of water available in a basin, the level of reserves in an oil field, the available stock of fish, the volume of timber inventories in a forest, and the levels of insect density are just a few situations where statistical inference is typically required to estimate stock levels from sample data. Although estimating the levels of the varying kinds of resources may require different sampling techniques, failure to utilize the complete knowledge base available for estimating the stock level results in suboptimal estimates. Furthermore, since estimation is an integral part of the decisionmaking problem, failure to obtain optimal estimates implies an overall solution that is suboptimal. If the view is taken that empirical natural resource problems are stochastic control problems, then the optimal estimate of the resource stocks and the optimal control actions generally must be determined simultaneously as discussed by Aoki.