Research Document - 2016/095
Fitting a statistical catch at length model (NFT-SCALE) to Unit 1 + 2 redfish (Sebastes mentella and Sebastes fasciatus)
By Duplisea, D.E., Bourdages, H., Brassard, C., Gauthier, J., Lambert, Y., Nitschke, P., and Valentin, A.
Abstract
A statistical catch at length model (NFT-SCALE) was applied separately to the two species of redfish in the Units 1 and 2 stock area. The SCALE approach is considered advantageous over a previously applied production modelling approach in that it incorporates composition data from both surveys and the commercial fishery and as such it has an ability to show relative strength of specific cohorts both and pre-recruit and adult sizes but it remains to be seen if this approach can be usefully applied to Units 1 and 2 redfish.
The SCALE fitting to these stocks captures many of features of the stocks including adult abundance index trends, recruitment index trends and relative cohort strengths; however, the fit to survey and commercial composition data is not as good:
- There is an antagonistic relationship between fitting the important indices of adult and recruit abundance and fitting composition but more so with Sebastes mentella than S. fasciatus. The model does however capture composition data better in years when the stock was more abundant (pre 1992) and marginally better in commercial catch than in the main survey (Unit 1).
- For S. mentella, weighting the Unit 1 adult index more heavily forced the biomass to capture high survey biomass in the 1980s, which increased residuals for catch at length (CAL) and increased the q estimate of the Unit 2 survey. For S. fasciatus, a smaller weight was needed on the adult index to better follow the 1980s high biomass and this did not have a strong effect on fit to composition data.
- SCALE tends to overestimate proportion of small S. mentellaand under estimate the proportion of larger S. mentella. Modifications of growth rate can change this to a small degree but not substantially but this composition misfit is less strong for S. fasciatus.
Retrospective runs from 2015-2005 for S. fasciatus show variation in the 1970s-1990s but settled down to a common trajectory after about 2006. When the U1 adult index was over-weighted (run2) the retrospective pattern was less apparent, especially after 1995.
The base run (run with fewest assumptions) for both species was quite good in capturing many of the stock features but better diagnostics and the prime directive to fit the over survey indices well (Francis 2011) led to preferred runs which weighted the adult abundance index for Unit 1 by a factor of between 3 and 7 in the likelihood function.
The fittings for both species showed that the 2011 cohorts are amongst the largest ever produced by these stocks since 1960 and it is anticipated that within 2-3 year the exploitable biomass of both stocks will be much larger than it has been in decades.
Statistical catch at length approaches hold promise for reproducing the dynamics of these redfish stocks. The main issues in accepting this class of model for these stocks concerns the subjective choice of what data to fit best or if new mechanisms need to be introduced to better explain some residuals.
Accessibility Notice
This document is available in PDF format. If the document is not accessible to you, please contact the Secretariat to obtain another appropriate format, such as regular print, large print, Braille or audio version.
- Date modified: