Data Prioritization for Stock Assessments in the Southeastern United States; How Much Do Better Data Actually Improve Assessment Accuracy?

“We need better data” is often the rally cry of stakeholders when asked for public comments about the status of a fishery stock.  With all the different data types and costs, it is difficult to discern the importance of each data type for assessment purposes.  We approached this problem using a simulation study.  We created an amalgam species from eight assessed stocks in the southeastern US and simulated a “known” population and assessment from that amalgamation.  We then incrementally improved the data, by either improving precision or sample size, for each data source, and groupings of data (e.g. all commercial data, all recreational data, or all survey data).  We also considered the marginal cost of each of these improvements.  Our results show that the age composition data have the most impact on the accuracy of our assessments.  Composition data are a relatively inexpensive type of data as well.  Within the age composition grouping, the survey age composition had the biggest impact on the assessment accuracy.  Further studies will examine other model constructs and incorporate the reduction of bias and other data sources we were unable to test in this study, such as the type of reproductive data.