Assessment of Indian Ocean Black Marlin (Makaira indica) Stock Using JABBA

Six scenarios were run using the Bayesian State-Space Surplus Production Model JABBA to assess the Indian Ocean black marlin (Makaira indica) based on alternative specifications of the Pella-Tomlinson model type that incorporated three differing CPUE input data series, three differing r priors and associated input values of B_MSY/K and two different values for process error. A general increase in black marlin catches is evident from 1990 onward with steep increases from 2010. Relative abundance (CPUE) trajectories show a steady decline from 1979 until 2005, after which signals of an increasing trend become apparent. The ‘drop one’ sensitivity analysis on CPUE indices indicates that omitting any index has a negligible influence on the management reference point estimates. However, the retrospective analysis produced an undesirable retrospective pattern as evident by systematic negative departures from the full model - the pattern becomes particularly strong from 2014 onward when the increase in total catch accelerated. Furthermore, an implausible trajectory is evident in all six Kobe biplots, which suggest that black marlin B/B_MSY increases with an associated increase in F/F_MSY for the period 2010-2016. These diagnostics highlight the poor performance with regard to the robustness of estimates and forward projections of B/B_MSY and F/F_MSY in this assessment and suggest model misspecification due to a simultaneous increase of both catch and CPUE, which is in conflict with the basic population dynamics principles. As such, resultant management reference points should be treated with caution.