synchSalmon
Analyses to explore drivers and consequences of component variability and synchrony using a combination of retrospective analyses and closed-loop simulations
Authors: C. Freshwater
*Date: March 14, 2019
Contact: cameron.freshwater@dfo-mpo.gc.ca
Summary
The synchSalmon repository contains the data and models necessary to complete the analysis within Weakened portfolio effects constrain management effectiveness for population aggregates (submitted). The goal of this study was two-fold. First, we conducted a retrospective analysis using time series of stock recruitment data from the Fraser River sockeye salmon stock aggregate to quantify changes in component variability (mean temporal coefficient of variation within a stock) and synchrony (covariance among stocks). Second, we used a closed-loop simulation model to evaluate how different levels of component variability and synchrony influence the probability of achieving a suite of management objectives. The closed-loop simulation model was developed to address generic questions related to rebuilding and is contained within a separate package samSim.
To install samSim
install.packages("devtools")
devtools::install_github("CamFreshwater/samSim")A summary of relevant files is provided below. Most functions are provided in samSim and contain relatively detailed documentation (and sometimes functioning examples). Details on how to run a simulation are provided in exampleSimRun.Rmd in the samSim package's repository under Rmd directory. Details on the operating model (biological dynamics and fishery interactions) and the management procedures (harvest control rule and assessment process) will be provided in the samSim vignette at a future date.
Note: development of samSim is ongoing and current package versions may no longer be directly compatible with the code provided here. Use the Git release associated with the published paper to guarantee compatibility.
Zenodo: 
Files
The repository contains the following directories:
data
Includes subdirectory:
- sox - contains stock-recruitment data necessary to complete retrospective analysis as well input .csv files used as inputs to the
recoverySim()function
figs
Contains figures (main text and supplementary) included in MANUSCRIPT TITLE.
outputs
Includes subdirectories:
- diagnostics - Automatically generated by
recoverySim()and used to store diagnostic plots representing the outputs from single trials. - generatedData - Used to store data files generated in post-analysis scripts.
- simData - Automatically generated by
recoverySim()and contains output from simulation runs. Each directory within simData represents a series of related scenarios, in this case each represents a given component variability and productivity operating model with multiple synchrony operating models nested inside.
scripts
Contains R scripts necessary to conduct retrospective and forward simulation analyses. Focal scripts are listed below, remainder are helper functions necessary to run Stan or TMB models.
- runModelRobustOM_synch_noSkew.R - Runs forward simulation and generates majority of figures contained in manuscript.
- runModelRobustOM_synchSensitivity.R - Runs forward simulation sensitivity analyses and generates relevant figures.
- supplementalFigures.R - Generates supplemental figures showing relevant probability distributions and an example of total allowable mortality harvest control rule.
- synchRetroAnalysis.R - Cleans stock-recruit data to pass to synchronyRetro-window-tmb.R and generates retrospective figures.
- synchRetroAnalysis_suppShortTS.R - Cleans stock-recruit data to pass to synchronyRetro-window-tmb.R and generates retrospective figures for retrospective sensitivity analysis.
- synchronyRetro-window-tmb.R - Estimates component variability, synchrony, and aggregate variability (with uncertainty) using a moving window.