Re-analysis of Virtual Population Analysis and State-Space Assessment Model for Operating Models of Chub Mackerel Stock Assessment in NPFC

The Technical Working Group for Chub Mackerel Stock Assessment (TWG CMSA) in NPFC has decided to use an operating model (OM) for comparing the performance of different four assessment model candidates. In this paper, we report the updated results of tuned virtual population analysis (VPA) and state-space assessment model (SAM), candidate stock assessment models proposed by Japan, under the determined scenarios to include biological uncertainties on natural mortality, weight, and maturity. The recent abundance estimates in VPA were much higher than those in the SAM and the recent fishing pressure was lower in the VPA. The scenarios under the highest maturity and weight estimated higher SSB in recent years and larger retrospective biases of SSB than the other scenarios in both SAM and VPA. The application of continuous hockey-stick stock-recruit model enabled to estimate feasible biological reference points. We found a few potential problems in both SAM and VPA: (1) the abundance indices from China and Russia were extremely hyper-stable against vulnerable stock size, (2) depletion statistics such as SSB/SSB0 and SSB/SSBmsy were highly sensitive to the choices of stock-recruit function, and (3) the MSY-based reference points were moderately sensitive to the biological parameters of maturity- and weight-at-age. How to treat these problems will be a matter of discussion in the TWG CMSA towards the construction of OM and the benchmark stock assessment.