On the development of acoustic descriptors for semi-demersal fish identification to support monitoring stocks

Abstract

Abstract Despite worldwide economic importance of lutjanid snappers, there are critical data gaps hindering accurate abundance estimates of these species, thereby impeding stock management efforts. Development of an active acoustics method could fill many of these data gaps, significantly benefiting stock assessment of snapper species that form typically mixed-species assemblages in non-trawlable regions. These methods are well suited for resource surveys but require the ability to derive accurate size estimates from target strength (TS) and the separation of echoes from species within the stock. This work evaluates the feasibility of active acoustics for surveying a Hawaiian bottomfish stock, referred to as the Deep-7 complex, consisting of six species of lutjanid snappers and a single grouper. Acoustic backscatter data and complementary methodologies were used to model TS and develop acoustic descriptors such as aggregation shape, size, and density, individual swimming pattern, distance from bottom, and bottom depth for Deep-7 species. Paired observations of in situ TS and fork length (FL) gave an estimated bulk relationship of 23.01·log10(FL)–72.18. The success of 25 accuracy tests indicates the effectiveness of the developed echo allocation parameters. Preliminary results imply that acoustics could be successfully used to monitor Deep-7 with broader applicability upon adapting methodology to other regions.