Improving data limited methods for assessing Indian Ocean neritic tuna species

[摘要] The Indian Ocean Tuna Commission (IOTC) manages 16 tuna and tuna-like species in the Indian Ocean, including six neritic tuna that are predominantly harvested along the coastal areas. The six species are: Bullet tuna (BLT, Auxis rochei), Spanish mackerel (COM, Scomberomorus commerson), Indo-Pacific King mackerel (GUT, Scomberomorus guttatus), Frigate tuna (FRI, Auxis thazard), Kawakawa tuna (KAW, Euthynnus affinis), and Longtail tuna (LOT, Thunnus tonggol). Unlike more valuable species, only limited data, primarily catch, exist for neritic tunas. So far it is difficult to conduct formal stock assessment for neritic tunas, and the data-poor, catch-based methods have been used instead. In recent years, development of data-poor methods is an active research area and new methods frequently appear in journal publications. The Commission recognized a need to review available data-limited methods that are potentially applicable for neritic tunas, and to incorporate new information in the catch-based methods currently used for neritic tunas.In this report, we conduct a literature review on data-limited methods. We categorize methods into traditional stock assessment, area-based ERA methods, age-based methods, length-based methods, and catch-only (or catch-based) methods. Catch-based methods have been adopted for neritic tuna assessment in the past several years and are deemed the best choice for the available data in IOTC. Besides catch-based methods, it is possible to use area-based methods, particularly the sustainability assessment for fishing effect (SAFE) approach, to assess fishing mortality status for neritic tuna because SAFE is flexible to accommodate varying data types and there appears to have sufficient information for such an analysis for several neritic tunas. Length-based methods require a range of assumptions that are difficult to meet for widely distributed migrating species that are captured by various gear types at different life stages, such as neritic tunas. However, because length data is the second most abundant information held by IOTC Secretariat, it would be interesting to explore length-based methods and see whether they can provide meaningful fishery status information.The second section of the report aims to improve the estimation of priors needed for catch-based methods:the level of stock depletion and the rate of intrinsic growth. We focus on two promising catch-based methods: the optimized catch-only method (OCOM) and Catch-MSY (CMSY), because the two approaches do not assume a fixed level of stock depletion as other catch-based methods do. The OCOM adopts a depletion prior using recently developed BRT (boosted regression trees) model, whereas CMSY derives depletion prior based on the ratio between last year’s catch and the maximum catch in the history of the stock. The two approaches also differ in deriving prior for the intrinsic population growth rate. For the OCOM, population growth rate is based on empirical correlation with other life-history parameters (mainly the natural mortality rate), whereas for CMSY it is based on a “resilience” parameter. Because data quality remains a concern, integrating the two approaches leads to a third method.The results indicate that Method 1 (OCOM prior) produces a higher r than Method 2 (CMSY prior) for all six species. The high r by OCOM prior causes a low K while a low r by CMSY prior causes a high K. The joint effect results in a similar MSY by either methods. Without independent study to compare the two methods, the integrated method may have an advantage over using either one of them.

[发布日期] 2019-12-30 [发布机构] CSIRO

[学科分类] 地球科学(综合)