Identifiable state-space assessment models: A case study of a fishery


  • Author: Yihao Yin, William H. Aeberhard, Stephen J. Smith and Joanna Mills Flemming
  • Date: 08 May 2019
  • Copyright: Image copyright of Patrick Rhodes

State‐space models (SSMs) are now popular tools in fisheries science for providing management advice when faced with noisy survey and commercial fishery data. Such models are often fitted within a Bayesian framework requiring both the specification of prior distributions for model parameters and simulation‐based approaches for inference. In a paper published in a special issue for The Canadian Journal of Statistics, the authors present a frequentist framework as a viable alternative and recommend using the Laplace approximation with automatic differentiation, as implemented in the R package Template Model Builder, for fast fitting and reliable inference. Additionally they highlight some identifiability issues associated with SSMs that fisheries scientists should be aware of and demonstrate how our modelling strategy surmounts these problems.

The paper is available via the link here and the authors explain their findings in further detail below:

Identifiable state‐space models: A case study of the Bay of Fundy sea scallop fishery

Yihao Yin, William H. Aeberhard, Stephen J. Smith and Joanna Mills Flemming

The Canadian Journal of Statistics, Volume 47, Issue 1

Special Issue: Special issue on Collaborative Research Team projects of the Canadian Statistical Sciences Institute, March 2019, pages 27-45

thumbnail image: Identifiable state-space assessment models: A case study of a fishery

Fish stock assessments are the foundation of fisheries management. They provide managers with science-based advice that is used to make regulatory decisions aimed at ensuring a sustainable fishing industry. A fish stock is a population of a certain species in a defined area. Stock assessment models attempt to integrate all available knowledge (e.g. growth rate, longevity, reported landings from commercial fisheries, indices of relative abundance from systematic surveys) to estimate important features of the stock, like its size, that are typically not directly measurable. State space assessment models (SSAMs), in particular, have become increasingly popular since their hierarchical structures help to distinguish different sources of randomness, which is essential when dealing with complex fisheries data.

One drawback of SSAMs is that model parameters may not all be identifiable or estimable based on the available data. The complexities of SSAMs used for contemporary fisheries management make these issues difficult to determine beforehand and also dependent on the estimation method. The Bayesian estimation framework assumes prior distributions for model parameters and estimates them by investigating posterior distributions. A frequentist (i.e. Maximum likelihood) framework does not require such prior knowledge and only relies on the information provided by the data summarized in the (marginal) likelihood. The estimates from both methods tend to coincide whenever the priors contain negligible information and identifiability and estimability issues are then diagnosable (typically via simulation). Unfortunately informative priors, set by default and/or based on arbitrary guidelines, can mask these important issues.

This paper explores SSAMs currently used for the highly valuable Bay of Fundy sea scallop fishery. The Bayesian version is shown to be problematic and an alternative frequentist model framework proposed. Through extensive simulations, this alternative formulation is established as estimable with available data. It is implemented using the R package Template Model Builder (TMB) to yield fast and reliable results. Moreover, it predicts lower population abundances than that of its predecessor which may translate into more conservative fishing advice going forward.

Related Topics

Related Publications

Related Content

Site Footer


This website is provided by John Wiley & Sons Limited, The Atrium, Southern Gate, Chichester, West Sussex PO19 8SQ (Company No: 00641132, VAT No: 376766987)

Published features on are checked for statistical accuracy by a panel from the European Network for Business and Industrial Statistics (ENBIS)   to whom Wiley and express their gratitude. This panel are: Ron Kenett, David Steinberg, Shirley Coleman, Irena Ograjenšek, Fabrizio Ruggeri, Rainer Göb, Philippe Castagliola, Xavier Tort-Martorell, Bart De Ketelaere, Antonio Pievatolo, Martina Vandebroek, Lance Mitchell, Gilbert Saporta, Helmut Waldl and Stelios Psarakis.