A stock assessment of Dissostichus eleginoides in the Northern Ross Sea, Antarctica

Abstract

This thesis conducts the first integrated stock assessment of Patagonian toothfish found in the Ross Sea Region, Antarctica, in order to estimate the stock size and status.

Patagonian toothfish are a large, long-lived, deep water species. Long-lived deep water species are characterised as being highly susceptible to over-exploitation. This has caused concerns over the impact of exploitation in toothfish fisheries among conservation parties. The Patagonian toothfish is currently on the Greenpeace red list, a list which deems the fisheries are not sustainable and should not be made commercially available.

The fishery in the Ross Sea region catches two Toothfish species, the larger more abundant Antarctic toothfish Dissostichus mawsoni and it’s sister species the Patagonian toothfish Dissostichus eleginoides. Both are regulated by the Commission for the Conservation of Antarctic Marine Living Resources (CCAMLR). CCAMLR is an international commission, which follows an ecosystem-based management approach. Fishing of Dissostichus spp. in the Ross Sea region is restricted to longline fishing only. Fishing of Dissostichus spp. in the Ross Sea region was initiated by a single vessel in 1997 and has been active ever since.

Data used included catches by the fishery, effort data from the fishery, and length-frequency data collected by scientific observers on board fishing vessels from the years 1997-2014. Data were characterised to identify any trends in catch, effort, depth, location, and timing of the fishery as well as biological features of the catch. The first four years of the fishery were different to the rest of the time series; a change in catch, catch depth and distribution, and number of vessels and fishing gear was apparent. However all data were included in the assessment as the Patagonian toothfish was a data-limited stock.

The assessment used an age-structured model, for the period 1997-2014, and fish ages 1-51. A single stock was assumed, with one area, and 50/50 sex ratio. Model variables included growth, maturity-at-age, vulnerability-at-age, natural and fishing mortality, and the stock-recruitment relationship. Observed data were catch, Catch Per Unit Effort (CPUE) as an assumed biomass index, and length-frequency of the fish catch. Some parameter estimates were assumed from an adjacent Patagonian toothfish fishery.

Parameters were estimated in the model using Bayesian methods, which produces a posterior distribution combining the likelihood of the observed data and a prior probability distribution for parameters. The posterior distribution was sampled using Markov Chain Monte Carlo (MCMC) methods and Gibbs sampler. Non-informative priors were mostly used. Sensitivity tests were completed to assess the impact of priors, and relative weighting of data.

The model led to two alternative hypotheses to fit observed data: (1) biological data were correct but catch-at-age in the beginning of the time series were wrong; or (2) the catch data were correct but the length-frequency data and/or biological parameters were incorrect. All model runs indicated the Patagonian toothfish stock had been heavily depleted during the fishery, with unfished spawning (reproductive) stock biomass (B₀) reduced to between 12% B₀ and 0.1% B₀ by 2014.

Discussion points include the method of parameter estimation, data availability, choice of priors, and data weighting. Future work could include collection of new biological data, investigating the reliability of CPUE and catch data, and investigating the use of alternative likelihoods and data weighting methods.

If the stock assessment conducted in this thesis is to be believed then the fish stock of Patagonian toothfish in the Ross Sea region has all but been wiped out, and concerns raised about the sustainability of the fishery and exploitation of the fish stock are shared here.