The SSC supports the CPT recommendation to bring forward a stock assessment based on GMACS for stock status and OFL/ABC determination in fall 2016. Specifically, the four GMACS models will be brought forward as alternatives with the GMACS match model replacing the 2015 model as the “status quo.”
The SSC offers the following comments to the stock assessment authors.
[ ] 1. Although standard procedure for many stock assessments on the West coast, the inclusion of extra CV seems to be rather arbitrary based on the numbers of points that fall within confidence intervals estimated from trawl surveys. Alternative approaches for including extra variability should be evaluated.
[ ] 2. The descriptions of seasons in the model are confusing and currently read as if M differs among seasons (see p. 39). More justification is needed on how seasons are defined and how they were selected, as well as clarification on M during these seasons.
[ ] 3. During the presentation to the SSC, uncertainty was expressed about the origins of the growth transition matrix, but page 7 of the report indicates that the matrix was derived by Otto and Cummiskey (1990). As this matrix is critical to the model, the origin and integrity of the growth transition matrix should be carefully explained in the assessment for fall 2016. In some other models, the transition matrix can be estimated. If there are doubts about the veracity of the transition matrix, perhaps this can be explored in the modeling framework.
[ ] 4. The selectivities were constrained so that they do not exceed 1.0, but the tables of log-transformed parameter estimates do not indicate that this upper bound was approached. This should be clarified.
[x] 5. It would be helpful to include a table of NMFS trawl survey CPUE by crab stage, just as was provided for the ADF&G pot survey (Table 1).
[ ] 6. Page 10 refers to a table of observed and estimated sample size, but no such table was provided.
[ ] 7. As with the 2015 model, GMACS consistently overestimates trawl survey estimates of male biomass in the last decade, whereas GMACS tends to underestimate the last couple of pot survey estimates (Figures 9, 12). This is also reflected in patterns in residuals, and the proportions of stage-3 crab tend to be overestimated in recent years (Figure 14). These patterns should be discussed in the assessment.
[ ] 8. The report contains very little description and interpretation of results. Moreover, not all figures are cited in the document. The document should highlight the major features of the results and offer some explanation, as well.
[ ] 9. A brief explanation was provided about the future outlook (page 12) that indicated a declining stock. However, stock trends shown in Figure 24 generally suggest population growth since 1993. Closer examination of Tables 9-11 suggest that trends depend somewhat on model run and life stage. Statements about future outlook should be qualified and refer to figures and tables and explain any differences in outcomes.
[ ] 10. The SSC discussed the possibility that these patterns could be indicative of spatial patterns in stock distribution. The trawl survey covers a much larger geographic distribution than the pot survey (Figure 4). Crab distribution may vary with sex (females tend to be found close to shore) and life stage. Thus, the trawl and pot surveys may sample the crab stock differentially. Moreover, the geographic distributions of these stages may vary with stock density and temperature. It could be informative to conduct some spatial analyses, which could include: (1) estimation of survey catchability as a function of temperature, (2) a stock assessment model run that includes pot surveys and only those trawl stations that fall within the pot survey distribution as a comparison to the runs that include the full trawl survey data, and (3) analysis of the spatial distribution of surveyed crabs by stage at high and low biomass and during warm and cold years.
[ ] 11. The CPT offered many insightful comments including recommendations on general code development for GMACS and the SMBKC application. The SSC appreciates and endorses the CPT recommendations.
The SSC supports the CPT recommendation to bring forward a stock assessment based on GMACS for stock status and OFL/ABC determination in fall 2016. Specifically, the four GMACS models will be brought forward as alternatives with the GMACS match model replacing the 2015 model as the “status quo.” The SSC offers the following comments to the stock assessment authors.