How to validate a length-based model of single-species fish stock dynamics

• Published in: Canadian Journal of Fisheries and Aquatic Sciences
• Main Authors: Kristensen, Kasper, Lewy, Peter, Beyer, Jan E
• Format: Article in Journal/Newspaper
• Language: English
• Published: Canadian Science Publishing 2006
• Subjects: atlantic cod; Gadus morhua
• Online Access: http://dx.doi.org/10.1139/f06-135; http://www.nrcresearchpress.com/doi/pdf/10.1139/f06-135

This paper validates a new length-based model of the dynamics of fish stocks or crustaceans by hierarchically testing statistical hypotheses and thereby investigating model complexity. The approach is based entirely on scientific survey data and on determination of the statistical distributions of the number of fish caught per haul in each length class. In our example, the negative binomial distribution is statistically accepted and linked to the population level through the new length-based model. The model is derived from the characteristics of continuous recruitment, individually based growth, and continuous, length-dependent mortality rates. Continuous recruitment with annually varying recruitment peaks and individually based growth was crucial for obtaining a model that could be statistically accepted. Natural mortality was estimated as well by the model. The model was applied to survey data for Atlantic cod (Gadus morhua) in the Baltic. Its simple generic nature, as well as the validation procedure, is useful in studying and understanding life history and stock dynamics.