Are stock assessment methods too complicated?

AJR Cotter, Mary Louise Burt, CGM Paxton, C Fernández, Stephen Terrence Buckland, J-X Pan

Research output: Contribution to journalArticlepeer-review

73 Citations (Scopus)

Abstract

This critical review argues that several methods for the estimation and prediction of numbers-at-age. fishing mortality coefficients F, and recruitment for a stock of fish are too hard to explain to customers (the fishing industry, managers, etc.) and do not pay enough attention to weaknesses in the supporting data, assumptions and theory. The review is linked to North Sea demersal stocks. First, weaknesses in the various types of data used in North Sea assessments are summarized, i.e. total landings, discards. commercial and research vessel abundance indices. age-length keys and natural mortality (M). A list of features that an ideal assessment should have is put forward as a basis for comparing different methods. The importance of independence and weighting when combining different types of data in an assessment is stressed. Assessment methods considered are Virtual Population Analysis, ad hoc tuning, extended survivors analysis (XSA). year-class curves, catch-at-age modelling, and state-space models fitted by Kalman filter or Bayesian methods. Year-class curves (not to be confused with 'catch-curves') are the favoured method because of their applicability to data sets separately, their visual appeal, simple statistical basis, minimal assumptions. the availability of confidence limits, and the ease with which estimates can be combined from different data sets after separate analyses. They do not estimate absolute stock numbers or F but neither do other methods unless M is accurately known. as is seldom true.

Original languageEnglish
Pages (from-to)235-254
Number of pages20
JournalFish and Fisheries
Volume5
Publication statusPublished - Sept 2004

Keywords

  • catch-at-age modelling
  • fisheries data
  • natural mortality
  • stock assessment.
  • virtual population analysis
  • year-class curves
  • AT-AGE DATA
  • NORTH-SEA PLAICE
  • IMPORTANCE RESAMPLING ALGORITHM
  • SEQUENTIAL POPULATION ANALYSIS
  • CURRENT MANAGEMENT ADVICE
  • PLEURONECTES-PLATESSA L.
  • NATURAL MORTALITY
  • TRAWL-SURVEY
  • MAXIMUM-LIKELIHOOD
  • MEASURING DISCARDS

Fingerprint

Dive into the research topics of 'Are stock assessment methods too complicated?'. Together they form a unique fingerprint.

Cite this