On Statistical Methods that Permit and Improve the Use of Simpler Model-Based Strike Limit Algorithms

The diierence between the dynamics of a population model in an Aboriginal Whaling Management Procedure (AWMP) and the dynamics of either the true stock or the common control program for testing AWMPs is a form of model misspeciication. The potential estimation bias introduced by this misspeciication provides the opportunity to empirically improve model-based strike limits. We show how to adjust the application of an existing AWMP in order to exploit new data sources or to use old data sources more appropriately. The result can be better strike limits without the need to design ever more complex population dynamics models. For example, age proportion data can be used with the AWMP introduced by Punt and Butterworth (1997) without redesigning their model to be age-stratiied. A simple example is presented to isolate the key aspects of the approach. For this example, the proposed approach provides better estimation through more exible use of all available data, particularly when the informativeness of portions of the data diier.