Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
5765466 | Fisheries Research | 2017 | 6 Pages |
Abstract
Natural mortality (M) is both a highly consequential and difficult process to estimate in stock assessments. Because of its correlation with other influential parameters, and the inadequacy of data to understand how it varies with time and ontogeny, estimates of M from integrated assessment models are often considered implausible. This has lead to the assumption that estimates of M are highly uncertain within integrated assessments and it is commonly fixed at values consistent with understanding about the stock's life history or derived from external analyses. Researchers recently used simulation analysis to challenge this assumption and provided evidence that M might be estimable under certain conditions. Our research builds upon those results by using the recently proposed age-structured production model diagnostic to help identify the conditions under which M might be estimable. This diagnostic aims to determine if changes in the scale and trend of stock abundance can be explained by catch alone, which is a key indicator of the presence of a production function. We apply the production model diagnostic to the same suit of assessments used in the aforementioned simulations to determine if a relationship between estimability of M and the presence of a production function can be identified. Statistical and subjective approaches to interpreting the production model diagnostic were developed with the aim of providing guidance on when M might be estimable. Statistical approaches to identifying the presence/absence of a production function did not outperform the subjective measure, but meaningful guidance about estimating M is still apparent. Our results provide more weight to the notion of M being estimable under certain conditions, and we provide guidance on identifying those conditions.
Related Topics
Life Sciences
Agricultural and Biological Sciences
Aquatic Science
Authors
Tim Sippel, Hui Hua Lee, Kevin Piner, Steven L.H. Teo,