Using Bayesian Age-Structured-Analysis (ASA) Model for Herring Population Dynamics in Prince William Sound, EVOS Herring Program

These data are part of the Herring Program of the Exxon Valdez Oil Spill Trustee Council, project numbers 13120111-Q, 14120111-Q, and 16120111-Q, which is a multi-faceted study to determine why herring populations in Prince William Sound remain depressed since the early 1990s. One of the main goals of the Herring Program is to improve the ability to predict herring populations through research and monitoring. This dataset involves the development of a modified version of the age-structure-analysis (ASA) model currently used by the Alaska Department of Fish and Game (ADF&G) to predict herring biomass levels in Prince William Sound. This modified model is a Bayesian ASA model that uses Markov chain Monte Carlo (MCMC) to generate posterior distributions for model parameters and output. The Bayesian ASA model is written for the Automatic Differentiation Model Builder (ADMB) programming framework (Fournier et al., 2012). This dataset includes a written description of the Bayesian model, model files for ADMB (.R, .ctl, TPL, .dat, .PIN formats), model output files as code (.R and .RData formats) and tabular data (.csv format), and graphical summaries (in .pdf format), including posterior predictive intervals, parameter distributions, and key model outputs. Anyone who wishes to run the Bayesian ASA must download ADMB (http://www.admb-project.org/downloads) and R (https://www.r-project.org/).