5. Interpret
5.3 Mechanics of interpretation
Overview Introduction Statistical Assessment Mechanics of Interpretation Finding Meaning in the Data Results and Next Steps
Interpretation for in-season management
For many projects, especially in areas with large commercial fisheries, the use of the data starts with an analysis and interpretation for in-season management. For more on that topic, click here.
Stock status
In every jurisdiction where salmon are used by people, the public and some government agency charged with the use of natural resources will want some kind of assessment of how the stock is doing from time to time. When these assessments appear as formal reports they are usually called stock status documents. For more information on this topic and for access to some examples, click here.
Associated techniques for the interpretation of stock assessment data
- Should credit Kent Crabtree
There are simply too many possibilities to extensively cover all of common inference tools. Also, the complexity of these tools makes a detailed discussion beyond the our scope. Common post-season techniques are described in one of several good textbooks, a sample of which are cited below. An overview of field techniques, such as the operation of sonars, is covered in a handbook distributed by the American Fisheries Society (Johnson et al. 2007). Click here for more information on some techniques for interpretation of stock assessment data, including (1) area under the curve, (2) methods for setting escapement goals, (3) trend analysis, and (4) using data from multiple sites with hierarchical models.
A conceptually different set of methods have been developed to deal with the confounding interactions that occur among simultaneously acting human activities and natural processes. For instance, as described in depth in Steps 1 and 2 in the sections on understanding mechanisms, confounding commonly occurs when, for instance, factors such as deteriorating climate-driven oceanographic conditions occur at the same time as continuing alteration of habitats along the migration route of juvenile and adult salmon. Attempts to determine the relative importance of these different mechanisms in causing observed changes in salmon indicators requires that various comparison groups be identified in space, time, or both. A rich literature on methods to conduct such studies has accumulated. We commend readers to some compilations of that information in Schmitt and Osenberg (1996), Schwarz (1998), Downes et al. (2002) and Roni (2005).
Simulation analyses to evaluate potential performance of sampling designs and the affiliated methods of data analysis
As Steps 1 and 2 (goals/objectives and designs of monitoring programs) illustrated, some designs are more appropriate than others for meeting particular objectives. The same can be said here with respect to methods of data analysis; statistical models used to estimate indicators of salmon conservation status such as abundance and time trend in abundance will likely vary in their statistical performance (bias and precision). Therefore, various combinations of objectives, indicators, sampling designs, and methods of analysis should be considered. We did this by conducting some empirically based simulation analyses to evaluate performance of several potential combinations of sampling designs and methods of data analysis in different situations.
Click here to download a PDF file of a brief document (Holt et al. 2010) that describes an analysis of trade-offs between sampling cost and ability to achieve monitoring objectives. This document also illustrates how empirically based simulation modelling can be used to choose the best sampling design or method of data analysis based on the monitoring objective and specific indicator used to reflect that objective.
Click here to download another PDF file, which is a brief document (Dorner et al. 2010) that describes a study that examined how alternative monitoring designs affect the ability to answer large-scale questions about the relative contributions of climate-driven factors to changes in salmon productivity in the North Pacific Ocean, as compared to potentially confounding anthropogenic impacts occurring on local scales.
NEXT: Finding meaning in the data | Go Back
