User menu
In May and June 2018 a collaborative study was conducted between national scientists with expertise in Japanese, Korean, Seychelles, and Taiwanese longline fleets, an independent scientist, and an IOTC scientist. The meetings addressed Terms of Reference covering several important issues related to yellowfin and albacore tuna CPUE indices in the Indian Ocean. The study was funded by the Indian Ocean Tuna Commission (IOTC).
Terms of Reference
1. Validate and improve current methods for developing indices of abundance for the main IOTC species.
2. Provide indices of abundance for selected IOTC species to be presented at the IOTC Working Parties in 2018.
3. Provide support and training to national scientists in their analyses of catch and effort data.
4. The analyses will consider data to be provided by key industrial fisheries operating in the Indian Ocean, including data from Japanese, Taiwanese, and Korean longline fleet.
5. Analyses will be carried out in a series of meetings scheduled during 2018. After preliminary discussions/meetings between the consultant and participating data providers, preparations will be carried out for each dataset and methods for CPUE standardization developed (or further elaborated upon), which will be followed by a joint CPUE meeting between all participating countries and the consultant.
Tasks will include the following, to the extent possible in the available time:
6. Work with the IOTC Stock Assessment Officer to coordinate meetings between data holders and the consultant.
7. Load, prepare, and check each dataset, given that data formats and pre-processing often change between years and data extracts, and important changes to fleets and reporting sometimes occur in new data.
8. Conduct the following analyses to improve CPUE methods and prepare indices:
o Apply cluster analyses or alternative methods for identifying targeting. Develop CPUE standardizations for main IOTC species using reliable data from each CPC, with priorities given to yellowfin and albacore tunas in 2018. Prepare separate indices for each fleet, and joint indices. Thoroughly check all code and results in order to validate the final standardized indices series.
o Explore alternative modelling and data transformation methods in order to normalise residuals and to accommodate strata with no zero catches.
o Explore residual patterns spatially and among clusters, fleets and vessels through time, and change models where necessary to address any problems identified.
o Apply methods for estimating relative regional weights, so as to apportion relative abundance among regions.
o Explore other distributions to improve model fit.
9. Document the analyses in accordance with the IOTC Guidelines for the presentation of CPUE standardisations and stock assessment models, adopted by the IOTC Scientific Committee in 2014; and to provide draft reports to the IOTC Secretariat no later than 60 days prior to the relevant IOTC Working Party meeting.
10. Undertake any additional analyses deemed relevant by the IOTC Working Parties, Scientific Committee, or IOTC Secretariat.
All work is subject to the agreement of the respective fisheries agencies to make the data available.
As in 2017, this document covers only the joint indices of abundance, describing their development for yellowfin and albacore tunas. Results are reported only for albacore tuna, with yellowfin tuna results presented previously in a separate document to the Working Party on Tropical Tunas.
Other issues are covered in related papers that describe the data preparation, cluster analyses, and individual indices for each fleet.
Data for the four fleets were standardized for each region to estimate indices of abundance. Indices were estimated using two approaches, delta lognormal and lognormal + constant, but the main approach was the delta lognormal. All models included the explanatory variables year-quarter and 5° cell as categorical variables, and a cubic spline on hooks as a covariate. Models included either a cubic spline fitted to hooks between floats or a categorical variable for cluster. Additional models were run that included both variables. Some models included vessel identity as a categorical variable. Models were run for the period 1952-1979 without vessel identity, for the later period 1979-2017 with vessel identity, and for the whole period 1952-2017 both with and without vessel identity.
Figures and tables are provided for each set of indices, including both quarterly and annual indices. Diagnostic plots are also presented.
