The 2012 assessment of striped marlin in the south-western Pacific Ocean, that updates the previous assessment done in 2006, was presented to SC8. Besides updating the input data to December 2011, the main developments to the inputs compared to the 2006 assessment included:

• Japanese longline catches for 1952-2011 were revised downwards by approximately 50%;
• Nine revised and new standardised CPUE time series (with temporal CVs) derived from:
  • Aggregate catch-effort data for Japanese and Taiwanese longline fisheries;
  • Operational catch-effort data for the Australian longline fishery;
  • Operational catch-effort data for the Australian and New Zealand recreational fisheries.
• Size composition data for the Australian recreational fishery.

The main developments to model structural assumptions were to: fix steepness at 0.8; fix growth at the newly published estimates; estimate spline selectivities for the main longline fisheries; estimate logistic selectivity for the Australian recreational fishery; include time-variant precision in fitting the model to standardized CPUE indices; and remove conflict among the CPUE indices by taking only the Japanese longline index in model area 2 as being representative for the Ref.case.

Model estimates of absolute biomass were sensitive to assumptions about selectivity and to conflicts among the standardized CPUE time series. The reference case model (Ref.case) estimated selectivity functions that decrease with age for the main longline fisheries that achieved the best fit to the size data. The CPUE time series for the Japanese longline fishery in area 2 was selected for fitting the Ref.case model because this time series was considered to be the most representative of changes in overall population relative abundance. Alternative options for selectivity assumptions and the CPUE time series included in the model fit were explored in sensitivity and structural uncertainty analyses, and are presented as the key model runs.

The main conclusions were:

• That current levels of catch are below MSY but are approaching the MSY calculated using the recent [low] levels of recruitment estimated for the last four decades;
• Apart from those model runs that assumed lower natural mortality or steepness, Fcurrent/FMSY was estimated to be lower than 1, and the median value for this ratio was estimated at 0.58, therefore, based on these results, overfishing is not occurring in the striped marlin stock;
• The structural uncertainty analysis indicates a 50% probability that SBcurrent is less than SBMSY, and 6 of the 10 key model runs indicate the ratio to be < 1. Together with these results and the recent trend in spawning biomass, it was concluded that striped marlin is approaching an overfished state.