|Document Type:||Journal Article|
|Title:||Random effect estimation of time-varying factors in Stock Synthesis|
|Author:||James T. Thorson, Allan C. Hicks, R. D. Methot|
|Journal:||ICES Journal of Marine Science|
|Keywords:||random effect, Bayesian, maximum likelihood, hierarchical model, time-varying parameter, mixed effect, Laplace approximation, stock assessment, penalized likelihood, recruitment,|
There is increasing recognition that biological processes such as fishery selectivity, natural mortality, and somatic growth vary over time, and that time-varying processes are important to include in marine population dynamics models. However, previous penalized-likelihood estimation approaches provide little guidance regarding the variance of time-varying processes, and random-effect approaches are computationally infeasible or not implemented for many models and software packages. We therefore show that existing stock assessment models and software can be used to calculate the Laplace approximation to the marginal likelihood of parameters representing variability over time, and use Stock Synthesis to show how this can be used to select an appropriate value for variation over time in stock assessment models. Using North Sea cod and Pacific hake models as case studies, we show that this method has little bias in estimating variances for simulated data. It also provides a similar estimate of hake recruitment (log-SD = 1.43) to Markov chain Monte Carlo (MCMC) methods (log-SD = 1.68), and estimates a non-trivial magnitude (log-SD = 0.07) of variation in growth for North Sea cod. We conclude by discussing the generality of the proposed method and by recommending future research regarding its performance relative to MCMC, particularly when estimating multiple variances simultaneously.
|Full Text URL:||http://icesjms.oxfordjournals.org/content/72/1/178.abstract?etoc|
|Theme:||Recovery, Rebuilding and Sustainability of Marine and Anadromous Species|
Characterize vital rates and other demographic parameters for key species, and develop and improve methods for predicting risk and viability/sustainability from population dynamics and demographic information.
Develop methods to use physiological and biological information to predict population-level processes.