Modeling uncertainty in fish population dynamics
Large uncertainties may exist in modeling various processes determining fisheries population dynamics. The uncertainties may come from various sources such as environmental variations (process errors), measurement errors, and model errors. In order to quantify the uncertainties, an understanding of...
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
Memorial University of Newfoundland
2004
|
Subjects: | |
Online Access: | https://research.library.mun.ca/10743/ https://research.library.mun.ca/10743/1/Jiao_Yan.pdf |
Summary: | Large uncertainties may exist in modeling various processes determining fisheries population dynamics. The uncertainties may come from various sources such as environmental variations (process errors), measurement errors, and model errors. In order to quantify the uncertainties, an understanding of the complex model error structure in the population dynamic models and how the model error structure affects the parameter estimation is important. In this study I evaluated and quantified the uncertainties in modeling various processes of fisheries population dynamics using Monte Carlo simulations and applied the proposed methods to Atlantic cod stocks. -- The generalized linear model approach, which can readily deal with different error structures, were used to identify suitable model error structure in stock-recruitment modeling, stock biomass modeling, and age-structure population modeling. A simulation study was developed to evaluate the influence of stock mixing on the collection of fish growth data and estimation of growth parameters. The recent status of the Atlantic cod fishery in Divisions 2J3KL was evaluated using a composite risk assessment method which calculates the total risk of overexploitation in the cod fishery. I considered the uncertainties in both biological reference point and current fishing mortality estimates. -- I recommend that the generalized linear model be used to identify appropriate model error structures in stock-recruitment modeling, stock biomass modeling, and age-structure population modeling. I also suggest that stock mixing be incorporated into stock assessment models to improve the estimation of growth parameters in stock assessment. Uncertainty in both management reference points and in indicator reference points should be considered in evaluating stock status using the proposed composite risk assessment method. |
---|