Optimal design in catch at age sampling

Project: Optimal design in catch at age sampling
Project Code: SAM2004/02
Start Date: 1 October 2004
Completion Date: 30 September 2005

Overall Objectives:

1. To develop techniques to improve stock assessments of harvested fish stocks.

Specific Objectives:

1. To determine the optimal design in sampling for catch-at-age data

Rationale:

General

Catch-at-age data are important inputs to the assessment of some of the most valuable New Zealand fish stocks. In designing a sampling program to collect such data a balance must be found between the precision desired in the middle of the age-frequency distribution and that desired for the tails of the distribution. There are currently no guidelines as to what is an optimal balance. For example, if we wished to estimate the middle of the distribution (where proportions at age typically exceed 0.1) with c.v.s of, say, 15%, what would be an appropriate target c.v. for the tails of the distribution (where proportions are less than 0.01)? This proposed project is aimed at answering questions like this. The answers will depend, to some extent, on the details of each assessment, but it should be possible to provide broad guidelines.

One outcome of this project would be a recommendation as to how the Ministry of Fisheries should specify its desired level of precision for catch-at-age data. Current practice is to specify a mean weighted c.v.. This is equivalent to requiring that S = Σa sa be less than a target value, where sa is the expected standard error of the estimated proportion at age a. This statistic implies a certain desired balance between precision in the middle and tails of the age distribution. If this balance is found not to be optimal it would be sensible to modify the statistic. For example, switching to S = (Σa sa²)0.5 would put less emphasis on precision in the tails of the distribution, whereas S = (Σa sa0.5)² would do the opposite.

Strategic Relevance

All the objectives in this project are consistent with the Fisheries Resources goal in the Ministry of Fisheries Strategic Research Directions document.

The percentage allocation for this project will be attributed to the following Fishstocks according to rule 9.2 of the Fisheries (Cost Recovery) Rules 2001:

• All these deepwater stocks: CDL, OEO, ORH, RIB
• All these middle depth stocks: BAR, FRO, GSH, GSP, HAK, HOK, LIN, SBW, SKI, SWA, WAR, WWA
• All these inshore stocks: BCO, BNS, BUT, BYX, ELE, FLA, GAR, GMU, GUR, HPB, JDO, MOK, RBY, RCO, SCH, SKA, SPD, SPE, SPO, SPR, STA, TAR, TRU, YEM

The project is estimated to cost between $0 – $50,000.