User menu
An abundance index for skipjack and juvenile yellowfin tuna from 1970 to 2016 has been developed from Maldives pole and line catch and effort data. Solutions for missing data were a random effects component used to account for missing mechanization information on the fleet 1974-1979 (Medley et al. 2017a) and the reconstruction of vessel length information using a vessel survival regression (described in Medley et al. 2017c). Fishing power effects related to vessel length are explained using a segmented regression that accounts for different classes of vessel. Both skipjack and yellowfin are combined into a single multivariate model, with skipjack catch rates standardized through a log-normal regression and yellowfin through a delta-lognormal regression. Additional fishing power effects which have not been recorded in the data have been estimated using subjective priors based on an expert meeting, and these could be included in the model. The model was fitted obtaining a MCMC approximation to the Bayes posterior for the abundance indices using Stan software. Remaining issues include poor estimation of catch rates for the smallest vessels and unaccounted for differences among landing atolls, as the reasons for these differences are not understood. Also, recent declines in logbook reporting rates are a cause for concern. All raw anonymized data and analysis code have been provided for full review.
