Estimation of rates of migration, exploitation and survival using tag recovery data for western Australian "salmon" (Arripis truttaceus : Arripidae :Percoidei)

The catch history and fisheries biology of Arripis truttaceus is described to outline the need for information on migration and survival in stock asse ssment. A model was developed to quantify annual rates of exploitation, sur vival and migration of juvenile and mature fish in nursery and adult areas using tag recovery data, The basic procedure for estimating these rates was to construct a functional model for recoveries by age and location, and to seek parameter values (using non-linear estimation techniques) for this mo del that provide a best least-squares fit to the recovery data. The sensiti vity of the parameter estimates to some potential effects of tagging, and o ther possible biases, was assessed by analysing data subsets from which tag groups had been omitted according to 14 scenarios of potential bias and ti me and place of tagging. Annual rates of exploitation, were estimated at 0. 71 +/- 0.015 for adult fish, and 0.17 +/- 0.006 for fully recruited juvenil es, Best estimates of annual rates of natural survival were 0.54 +/- 0.043 for mature fish and 0.74 +/- 0.024 for juvenile fish. For maturing fish in the South Australian nursery area, the best estimates of annual rates of mi gration to the spawning area were about 0.2 for 4 year old fish, and 0.49 /- 0.046 for fish of five years and older. The relative vulnerability to fi shing in the South Australian nursery area was 0.51 +/- 0.047 for 0+ fish a nd 0.88 +/- 0.044 for 1 year old fish. Parameter estimates were most sensit ive to the omission of data from tagging of older fish in the nursery area, and on mature fish in the adult area. Delay of migration by the tagging pr ocedure caused underestimation of migration-at-ae. Lack of mixing of tagged and untagged fish was considered to cause overestimation of the relative v ulnerability to fishing of 0+ fish. Overall parameter estimates generally a greed with ancillary data from catch sampling and population surveys, but m igration rates:from nursery areas to the adult area were lower than expecte d. The best estimates provide the basis for other stock assessments to eval uate alternative harvesting policies in the fishery for Arripis truttaceus. (C) 2000 Elsevier Science B.V. All rights reserved.