Specify the stock-recruit relationship (SRR) for Rceattle

Stock recruitment relationships currently implemented in Rceattle:

srr_fun = 0 or "mean": No stock recruit relationship. Recruitment is a function of R0 and annual deviates (i.e. steepness = 0.99). $$R_y = exp(R0 + R_{dev,y})$$
srr_fun = 2 or "BevertonHolt": Beverton-holt stock-recruitment relationship $$R_y = \frac{\alpha_{srr} * SB_{y-minage}}{1+\beta_{srr} * SB_{y-minage}}$$
srr_fun = 4 or "Ricker": Ricker stock-recruitment relationship $$R_y = \alpha_{srr} * SB_{y-minage} * exp(-\beta_{srr} * SB_{y-minage})$$

When srr_pred_fun > 0 and srr_fun = 0 recruitment in the hindcast is estimated as in srr_fun = 0 $$R_y = exp(R0 + R_{dev,y})$$, but an additional stock recruitment relationship defined by srr_pred_fun is estimated between srr_hat_styr and srr_hat_endyr and treated as an additional penalty. The stock recruitment relationship defined by srr_pred_fun is then used in the projection.

Usage

build_srr(
  srr_fun = 0,
  srr_pred_fun = srr_fun,
  proj_mean_rec = TRUE,
  srr_mse_switchyr = NULL,
  srr_hat_styr = NULL,
  srr_hat_endyr = NULL,
  srr_est_mode = 1,
  srr_prior = 4,
  srr_prior_sd = 1,
  srr_indices = NA,
  Bmsy_lim = NA,
  linkages = NULL
)

Arguments

srr_fun: Stock recruit function to be used for hindcast estimation of Rceattle (see @description below). Default = 0
srr_pred_fun: stock recruit function for projection, reference points, and penalties to be used for Rceattle (see below). When srr_fun == 0, it treats the stock-recruit curve as an additional penalty onto the annualy estimated recruitment from the hindcast (sensu AMAK and Jim Ianelli's pollock model). If srr_fun > 0 then srr_pred_fun = srr_fun and no additional penalty is included.
proj_mean_rec: Project the model using: 0 = mean recruitment (average R of hindcast) or 1 = SRR(omega, srr_devs)
srr_mse_switchyr: is used for MSEs to deal with AMAK and Jim Ianelli's estimation where a stock recruit function is estimated as an additional penalty (srr_fun = 0 and srr_pred_fun > 0). It tells the model in what year to switch to the stock recruit function.
srr_hat_styr: Integer. The first year used for estimating the recruitment function as an additional penalty. It will add additional penalties sensu AMAK and Jim Ianelli's pollock model when srr_pred_fun > 0 and srr_fun = 0, starting at styr + 1. Defaults to $styr + 1$ in $data_list$. Useful if environmental data used to condition stock-recruit relationships is not available until end-year, but projections are desired.
srr_hat_endyr: Integer. The last year used for estimating the recruitment function as an additional penalty. It will add an additional penalties sensu AMAK and Jim Ianelli's pollock model when srr_pred_fun > 0 and srr_fun = 0. Recruitment Defaults to $endyr$ in $data_list$. Useful if environmental data used to condition stock-recruit relationships is not available for the full time-series, but projections are desired.
srr_est_mode: Switch to determine estimation mode. 0 = fix alpha to prior mean, 1 = freely estimate R0, alpha, and/or beta (default), 2 = use lognormally distributed prior for alpha (Ricker) or steepness (Beverton), 3 = use beta distributed prior for steepness (Beverton) given mean and sd.
srr_prior: mean for normally distributed prior for stock-recruit parameter
srr_prior_sd: Prior standard deviation for stock-recruit parameter
srr_indices: Soft-deprecated. Use the linkages argument instead. See vignette("environmental-linkages-and-priors").
Bmsy_lim: Upper limit for Ricker based SSB-MSY (e.g 1/Beta). Will add a likelihood penalty if beta is estimated above this limit. Default NA is not used.
linkages: Optional named list of linkage_spec() objects keyed by recruitment parameter name (must be one of "R0", "alpha", "beta"). Each spec describes how that parameter depends on environmental covariates and on stratifying factors (species, sex). The offset enters additively (on the log scale) inside the recruitment compute. See vignette("environmental-linkages-and-priors") for details.

Value

A list containing the stock recruitment relationship settings