Community Research and Development Information Service - CORDIS

Final Report Summary - ECOKNOWS (Effective use of ecosystem and biological knowledge in fisheries)

Executive Summary:
The ECOKNOWS project produced 36 peer-reviewed published papers and 23 articles published in non-scientific press. The project reviewed and improved model assumptions in fisheries, including e.g. features of stock-recruit models, natural mortality and species biology.

The project provided tools that improve the credibility of fisheries science by
1) using probability distributions in historical data analysis and predictions;
2) including ecological and genetic knowledge in fisheries models; and
3) widely utilizing available biological and ecological data-sets.

Methods were developed that utilize all relevant information for short-term probabilistic prediction. The methodology used survey data sets and existing large, currently poorly utilized, effort data sets (log book data) in ICES. For more effective learning, systematic use of prior information from data bases was developed, in which each additional data-set updates knowledge in sequential analysis. The use of priors and full probabilistic models were made more operational for European stock assessments. Moreover, a manual of best practices in deriving priors was prepared and offered for publication. The computational efficiency of models was improved mainly by novel solutions for parametrizations.

The innovative methodology developed in the project was applied to and demonstrated in the case studies of a) the European Anchovy in the southern Atlantic; b) Atlantic and Baltic salmon stocks c) clupeoids in the Baltic and North Sea; d) Baltic and Mediterranean multi-species fisheries, for which length-based models were applied ; e) Hake (Merluccius merluccius) in Division IIIa, Subareas IV, VI, and VII, and Divisions VIIIa,b,d (Northern stock); and f) Northern shrimp (Pandalus borealis) in Division IIIa and Division IVa East (Skagerrak and Norwegian Deep).

ECOKNOWS established implemented a Bayesian learning process into FishBase ( databases. The inference model developed to use the FishBase information provides estimates for key biological parameters and includes direct assessment of uncertainty of these estimates.

The philosophy related to new ways of using scientific information to stakeholders was disseminated and demonstrated in various occasions. An input to academic teaching material was provided and linked to FishBase to provide continuation of the learning after the project.

Project Context and Objectives:
Most of the currently applied European stock assessment methods rely on assumptions that contradict the basic understanding about ecological characteristics of fish life-cycle. For example, the natural mortality (M) is often assumed to be exactly known and constant over age groups and time, and independent of the population density and external factors such as predation and food availability. Similarly, variation and uncertainty in growth and maturation rates is often ignored. It is also not uncommon to assume that number of offspring would be independent of the amount of spawning stock biomass and spawned eggs when starting forward predictions. Assumptions like these contradict biological knowledge, ignore uncertainties about important biological processes and undermine the credibility of fisheries science. Traditional methodology cannot accommodate models that are justified but can be conditioned only on a data set that is small relative to realistic model’s complexity.
ECOKNOWS project applied the Bayesian approach in fisheries science and developed modelling tools which have more realistic assumptions than traditional assessment methods. Bayesian approach offers a sound framework for including information from multiple sources: ecological knowledge, experimental data, knowledge from similar populations, knowledge from published papers, and observed fishery data. Assessment of uncertainty, which is inherent to Bayesian approach, supports the applicability of precautionary approach. Effective use of prior knowledge improves the prediction capacity and the credibility of fisheries science.

Objectives of ECOKNOWS were
a) To critically review and improve current model assumptions in fisheries by using current knowledge in ecology.
b) To provide tools that improve the credibility of fisheries science by: 1) using Bayesian inference 2) by including ecological and genetic knowledge in fisheries models; and 3) by using all available biological and ecological data to estimate the probabilistic stock-recruit relationships.
c) To make use of priors and full probabilistic models as operational part of some European stock assessments.
d) To improve the computational efficiency of Bayesian models so they can be more effectively applied, e.g. in fish stock assessment working groups.
e) To apply the methods to European case studies, and to learn effectively from contrasts between areas and between stocks.
f) To establish Bayesian learning algorithms (prior–data analysis–posterior) to existing FishBase ( database to effectively support the risk assessment in fisheries.
g) To disseminate the philosophy related to new ways of using scientific information to stakeholders and to academic teaching material, and especially link this to common logic to improve the credibility of fish stock assessment.

Project Results:
The project developed and tested a general population dynamics model (GPDM) for fisheries stock assessment. The implementation of the model was very time-consuming, but given enough time, was successful in all case studies. The project also developed the first version of a learning database for biological information, and the further development of this database will continue after the project.
ECOKNOWS successfully reviewed and improved current model assumptions in fisheries. Assumptions considered, for example, reproduction dynamics and natural mortality. The project succeeded to provide tools that can improve the credibility of fisheries science by including ecological and genetic knowledge in fisheries models and by using available biological and ecological data to estimate the probabilistic stock-recruitment relationships.
ECOKNOWS applied the methods to the case studies and implemented a Bayesian learning algorithm to existing FishBase ( The posterior probability distribution of one analysis can be potentially used as a prior for the next one. Thus, the database supports the systematic use of Bayesian models in fisheries stock assessment. Moreover, basic biological knowledge can be used to decrease the uncertainty in population specific variables of interest.

Potential Impact:
The ECOKNOWS results have the following potential impacts:
1. The general population dynamics model (GPDM) can be adopted by fisheries stock assessment working groups worldwide. This is expected to increase the use of existing biological information while realistically accounting for the remaining uncertainty about the current and future states of the fish stock.
2. Accounting for uncertainty in fisheries management can help to avoid overfishing and consequent socioeconomic losses.
3. Use of more realistic assumptions in fisheries stock assessment is expected to enhance the credibility of the assessment results and consequently to enhance the commitment of stakeholders to management actions based on those results.
The findings and the philosophy of the project was widely disseminated to both scientific audience and stakeholders via publications in scientific and professional journals. The final symposium in June 2014 attracted a large international audience. The models and material produced during the project has been made publicly available also for educational purposes in FishBase. One of the most influential project dissemination activities is probably the ICES manual for the best practices on the use of prior information in fisheries stock assessment.

List of Websites:

Professor Sakari Kuikka
University of Helsinki
Department of Environmental Sciences
Viikinkaari 2 a
00790 Helsinki

Related information

Reported by

Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top