In Support of a Rationally Managed Fishery: Age and Growth in Patagonian Toothfish ( Dissostichus eleginoides )

• Main Author: Ashford, Julian R.
• Format: Doctoral or Postdoctoral Thesis
• Language: unknown
• Published: ODU Digital Commons 2001
• Subjects: Age; Chilean sea bass; Dissostichus eleginoides; Fishery; Patagonian toothfish; Southern Ocean; Aquaculture and Fisheries; Ecology and Evolutionary Biology; Kerguelen; Heard Island; Kerguelen Islands; Indian; Patagonian Toothfish
• Online Access: https://digitalcommons.odu.edu/biology_etds/45; https://digitalcommons.odu.edu/cgi/viewcontent.cgi?article=1049&context=biology_etds

Patagonian toothfish (Dissostichus eleginoides) occur on the continental shelves and shelf breaks of southern South America and the Southern Ocean. Stock structure, critical to good fisheries management, can be inferred from growth differences between areas, but available growth data are compromised by inconsistencies in age estimation methods, sampling and sample sizes, and techniques used to derive estimates. I asked the scientific question: how is growth in Patagonian toothfish structured spatially within the Southern Ocean? I developed a multi-stage randomized design to sample fish caught by commercial longline, and an age estimation methodology. Because toothfish are difficult to age, I developed an ANOVA model for estimating precision and accuracy of age data relative to a standard, as the basis for a protocol for quality control of age data. The methodology was used to obtain age data from toothfish taken from the Falkland Islands and South Georgia in the South Atlantic, and the Kerguelen Islands and Heard Island in the southern Indian Ocean. I estimated von Bertalanffy growth parameters for each area, constructed models to describe rival hypotheses of stock mixing and separation between areas, and selected between the models using normal likelihood methods. The abundance of the captured population varied at a scale of c500 m (76% of variance), and between fishing days (24%). Most variation in length composition was captured at scales less than 500 m (79%). I calculated that sampling 16 10-coil lengths of line/day on 36 days of a voyage of 60 days hauling would be the optimal sampling strategy. Significant bias in age estimation was found between readers and between readings by one reader but, once accounted for, precision of age estimation remained similar between sexes; however, a validation test of the accuracy of the age estimation methodology was inconclusive. Growth data supported the hypothesis of stock separation between the Falkland Islands and South Georgia, but not between South Georgia and ...