Application of three data-limited methods for stock assessment of neritic tunas and mackerels in the Indian Ocean

Abstract:

Neritic tuna and mackerels catch in the Indian Ocean have increased rapidly in recent years, whereas these species currently lack comprehensive assessment due to their predominant capture in artisanal or small-scale fisheries, where fishery statistics are insufficient and the necessary data required for routine stock assessment are lacking. In order to better understand the resource status and develop appropriate management measures, this study applied three data-limited methods （Monte carlo catch-msy,Depletion-based stock reduction analysis, Optimized catch-only assessment method） to assess Bullet tuna, Frigate tuna, Kawakawa, Longtail tuna, Indo-Pacific King Mackerel, and narrow-barred Spanish mackerel in the Indian Ocean. Stock status were evaluated based on relative biomass （B/B_MSY） and relative fishing mortality （F/F_MSY）. Results showed that Frigate tuna and Indo-Pacific King Mackerel had a healthy status [P（B/B_MSY>1）=78%,P（F/F_MSY<1）=67%； P（B/B_MSY>1）=78%, P（F/F_MSY<1）=55%], Bullet tuna and Kawakawa were at higher risk of overfishing [P（B/B_MSY>1）=78%,P（F/F_MSY<1）=33%；P（B/B_MSY>1）=78%,P（F/F_MSY<1）=45%]. Longtail tuna and narrow-barred Spanish mackerel were at higher risk of being overfished and were subject to overfishing [P（B/B_MSY>1）=33%,P（F/F_MSY<1）=44%；P（B/B_MSY>1）=55%,P（F/F_MSY<1）=33%]. Of the three models, CMSY and DB-SRA resulted in close MSYestimates, with CMSY giving the most cautious assessment results （overfishing in all six species） and OCOM giving the most optimistic results, with some differences in the judgement of the current resource status of the species among the three models. Sensitivity analyses showed that both the priori setting of r and B_end/K had a large impact on the CMSY results； DB-SRA was sensitive to B_t/K and more robust to K. All three models were applicable for neritic tunas and mackerels stock assessment； however, relying on a single approach may lead to bias results. It is recommended that when using data-limited methods, multiple models should be used to reduce bias. The findings of this study can provide a reference for the management of neritic fisheries in the Indian Ocean.