Package 'SPoRC'

Initial guess of F

Provided catch values

Numbers, dimensioned by ages, and sexes

Weight, dimensioned by ages and sexes

Natural mortality dimensioned by ages and sex

Fishery selectivity, dimesnioned by ages and sex

Number of iterations for bisection

Lower bound of F

Upper bound of F

Integer. Number of years to project in the closed loop.

Integer. Number of simulation replicates.

List. Observed data and configuration for the population.

List. Parameter values for the model.

List. Mapping of parameters for optimization.

List. Standard deviation reports from fitted model.

List. Report from fitted model.

Character vector of random effects estimated

Character or numeric. Fill method for fishery index standard errors when extending to simulation years. Options are: - '"zeros"': fill with zeros - '"last"': repeat last non-NA slice - '"mean"': fill with the mean of the observed series - Numeric: constant scalar or array value

Character or numeric. Fill method for survey index standard errors. Same options as 'FishIdx_SE_fill'.

Character or numeric. Fill method for fishery age composition input sample sizes. Options are: - '"zeros"', '"last"', '"mean"' (as above) - '"F_pattern"': fill based on fishing mortality pattern in the closed-loop simulation - Numeric: constant scalar or array value

Character or numeric. Fill method for fishery length composition input sample sizes. Same options as 'ISS_FishAgeComps_fill'.

Character or numeric. Fill method for survey age composition input sample sizes. Options are '"zeros"', '"last"', '"mean"', or a numeric constant.

Character or numeric. Fill method for survey length composition input sample sizes. Same options as 'ISS_SrvAgeComps_fill'.

Optional named arguments for custom inputs. Each name must correspond to a component expected by the simulation setup functions, and be dimensioned appropriately: 'Setup_Sim_Fishing()', 'Setup_Sim_Survey()', 'Setup_Sim_Biologicals()', 'Setup_Sim_Rec()', and 'Setup_Sim_Tagging()'. Examples include:

- **Fishing**: 'fish_sel_input', 'ln_sigmaC', 'Fmort_input', 'fish_q_input', etc. - **Survey**: 'srv_sel_input', 'srv_q_input', 'ObsSrvIdx_SE', etc. - **Biologicals**: 'WAA_input', 'MatAA_input', 'natmort_input', 'AgeingError_input', 'SizeAgeTrans_input' - **Recruitment inputs**: - 'R0_input', 'h_input', 'sexratio_input', 'ln_InitDevs_input', 'Rec_input' - 'Rec_input': - If shorter than the number of projection years, new recruitment deviates will be simulated based on 'recruitment_opt', 'R0_input', and 'h_input' (supports changing regimes across years). - If you want fixed recruitment for all projection years, provide a 'Rec_input' array that spans all years. - **Tagging**: 'Tag_Reporting_input', 'ln_Init_Tag_Mort', 'ln_Tag_Shed', 'tag_selex', 'tag_natmort' - **Movement**: 'Movement' (must match the expected dimensions and be named exactly 'Movement')

The values must have the correct dimensions expected by each component. If a component is not provided, default behavior will extend the last year (or zeros for fishing mortality, which can filled in subsequently)

List of data inputs

Report file list

Age labels

Length labels

Year labels

Data list to make obj

Parameter list to make obj

Mapping list to make obj

Character of random effects

sd for jitter (additive)

Number of jitters to do

Number of newton loops to do

Whether to do paralleizaiton or not (boolean)

Number of cores to use

Vector of parameter starting values to use for jitter analysis. The default of this is NULL (jitters the starting value of the model). If a vector is provided, the jitter is initialized at the MLE parameters

data list from model

parameter list from model

mapping list from model

character vector of random effects to estimate

parameter name we want to profile

Index for an parameter array, pointing to the value we want to map off (index is relative to a flattened array)

minimum value of profile

maximum value of profile

increment value between min and max value

logical, whether to use parallel processing (default FALSE)

integer, number of cores to use for parallel processing (default is detectCores() - 1)

Number of projection years

Number of regions

Number of ages

Number of sexes

Array of recruitment sex ratio (n_regions, n_proj_yrs, n_sexes)

Number of fishery fleets

Whether recruits move (0 == don't move, 1 == move)

Recruitment matrix dimensioned by n_regions, and n_yrs that we want to summarize across, or condition our projection on

Terminal Numbers at Age dimensioned by n_regions, n_ages, n_sexes

Terminal Unfished Numbers at Age dimensioned by n_regions, n_ages, n_sexes

Terminal fishing mortality rate, dimensioned by n_regions, n_fish_fleets

Natural mortality, dimensioned by n_regions, n_proj_yrs, n_ages, n_sexes

Weight at age, dimensioned by n_regions, n_proj_yrs, n_ages, n_sexes

Weight at age for the fishery, dimensioned by n_regions, n_proj_yrs, n_ages, n_sexes, n_fish_fleets

Maturity at age, dimensioned by n_regions, n_proj_yrs, n_ages, n_sexes

Fishery selectivity, dimensioned by n_regions, n_proj_yrs, n_ages, n_sexes, n_fish_fleets

Movement, dimensioned by n_regions, n_regions, n_proj_yrs, n_ages, n_sexes

Fishing mortality reference point dimensioned by n_regions and n_proj_yrs

Biological reference point dimensioned by n_regions and n_proj_yrs

Function describing a harvest control rule. The function should always have the following arguments: x, which represents SSB, frp, which takes inputs of fishery reference points, and brp, which takes inputs of biological reference points. Any additional arguments should be specified with defaults or hard coded / fixed within the function.

Recruitment simulation option, where options are "inv_gauss", which simulates future recruitment based on the the recruitment values supplied using an inverse gaussian distribution, "mean_rec", which takes the mean of the recruitment values supplied for a given region, and "zero", which assumes that future recruitment does not occur

Fishing Mortality option, which includes "HCR", which modifies the F reference point using a user supplied HCR_function, or "Input", which uses projected F values supplied by the user.

Fraction time of spawning used to compute projected SSB

A list object containing the following arguments:

recruitment_dd

A value (0 or 1) indicating global (1) or local density dependence (0). In the case of a single region model, either local or global will give the same results

rec_lag

A value indicating the number of years lagged that a given year's SSB produces recruits

R0

The virgin recruitment parameter

Rec_Prop

Recruitment apportionment values. In a single region model, this should be set at a value of 1. Dimensioned by n_regions

h

Steepness values for the stock recruitment curve. Dimensioned by n_regions

WAA

A weight-at-age array dimensioned by n_regions, n_ages, and n_sexes, where the reference year should utilize values from the first year

MatAA

A maturity at age array dimensioned by n_regions, n_ages, and n_sexes, where the reference year should utilize values from the first year

natmort

A natural mortality at age array dimensioned by n_regions, n_ages, and n_sexes, where the reference year should utilize values from the first year

SSB

All SSB values estimated from a given model, dimensioned by n_regions and n_yrs

Integer. Number of retrospective peels to perform.

List. Data input for the RTMB model.

List. Parameter values for the RTMB model.

List. Mapping information for the RTMB model.

Character vector. Names of random effects in the model. Default is NULL.

Logical. Whether to run retrospective peels in parallel. Default is FALSE.

Integer. Number of cores to use for parallel execution if do_par = TRUE.

Integer. Number of Newton loops to run during model fitting. Default is 3.

Logical. Whether to apply Francis reweighting within each retrospective peel. Default is FALSE.

Integer. Number of Francis reweighting iterations. Required if do_francis = TRUE.

List. Control parameters passed to nlminb during model fitting. Default is list(iter.max = 1e5, eval.max = 1e5, rel.tol = 1e-15).

Logical. Whether to return standard errors from sdreport. Default is FALSE.

Integer array. Lags for fishery index data [regions x fleets]. Default is zeros.

Integer array. Lags for fishery age composition data [regions x fleets]. Default is zeros.

Integer array. Lags for fishery length composition data [regions x fleets]. Default is zeros.

Integer array. Lags for survey index data [regions x fleets]. Default is zeros.

Integer array. Lags for survey age composition data [regions x fleets]. Default is zeros.

Integer array. Lags for survey length composition data [regions x fleets]. Default is zeros.

Integer. Lag for tagging data. Default is 0.

Vector of residuals

Whether to use mean 0 assumption of mean of residuals (default = use mean 0)

Type of test to do, less = left tailed test that detects positive autocorrelation, two.sided = two sided test that tests whether there is positive and/or negative autocorrealtion. The null is that there isn't any, rejecting the null (<0.05) indictes that there is some non-randomness.

Data list

Parameter list

Mapping list

Character of random effects to integrate out

Number of newton loops to run to get gradients down

Boolean on whether or not model run is silent

Boolean on whether or not model is optimized

List argument controls by nlminb

Additional arguments taken by MakeADFun

List of n_models of 'SPoRC' data lists

List of n_models of 'SPoRC' report lists

Vector of model names

List of n_models of 'SPoRC' data lists

List of n_models of 'SPoRC' report lists

Vector of model names

Logical. If TRUE, uses values from the simulation environment ('sim_env') for calculating reference points. If FALSE, uses 'asmt_data' and 'asmt_rep'.

Simulation environment

Optional list. Assessment data object (from RTMB) if not using true values.

Optional list. Assessment report object (from RTMB) if not using true values.

Integer. Number of years to include in calculations (usually the last year of the assessment or simulation).

Integer. Index of the simulation replicate in 'sim_env'.

List. Options for reference point calculations:

n_avg_yrs

Number of years to average over demographic rates. Default is 1.

SPR_x

Target SPR fraction for reference point calculations. Default is 0.4.

calc_rec_st_yr

Year to start calculating mean recruitment. Default is 1.

rec_age

Age at recruitment. Default is 1.

type

Reference point type: "single_region" or "multi_region". Default is "single_region".

what

Method for reference point calculation. Options include "SPR", "BH_MSY", "independent_SPR", "independent_BH_MSY", "global_SPR", "global_BH_MSY". Default is "SPR".

Number of projection years

list of data inputs

report file from RTMB

vector of observed age labels in assessment

vector of length labels in assessment

vector of years

List of n_models of 'SPoRC' data lists

Character vector of model names

Data list fed into RTMB

Report list output from RTMB

Year labels to use (vector)

List of n_models of 'SPoRC' data lists

List of n_models of 'SPoRC' report lists

Vector of model names

A list of model input data objects, one for each model (i.e., a list of SPoRC-formatted data lists). Each element should contain information on regions, years, ages, fleets, and biological inputs (e.g., weight-at-age, maturity, mortality).

A list of model output objects, one for each model (i.e., a list of SPoRC-formatted report lists). Each element must include recruitment, selectivity, mortality, and numbers-at-age.

A named list specifying options for reference point calculations. See Get_Reference_Points for more details. Must include:

SPR_x

Spawning potential ratio (e.g., 0.4) for calculating F reference points. May be NULL if using bh_rec.

t_spwn

Fraction of year when spawning occurs (e.g., 0.5).

sex_ratio_f

Proportion of recruits that are female.

calc_rec_st_yr

Start year for averaging recruitment.

rec_age

Recruitment age.

type

Reference point calculation method (e.g., "multi_region").

what

Type of output requested from the reference point function.

A named list of projection settings. See Do_Population_Projection for details. Must include:

n_proj_yrs

Number of years to project forward.

HCR_function

Harvest control rule function to use.

recruitment_opt

Recruitment assumption (e.g., "mean_rec", "bh_rec", "inv_gauss").

fmort_opt

Fishing mortality assumption (e.g., "input", "HCR").

n_avg_yrs

Number of years to average over for projection inputs.

A character vector of model identifiers (e.g., c("Base", "Alt1", "Alt2")), one for each element in data and rep.

Integer specifying the logistic-normal correlation structure:

• 2 = iid across categories

• 3 = AR1 across categories

• 4 = AR1 across categories and sexes

Number of composition categories (e.g., age or length bins). For comp_like = 2, 3, the covariance matrix is dimensioned n_bins. For comp_like = 4, it is dimensioned n_bins * n_sexes.

Number of sexes. Required when comp_like = 4.

Standard deviation parameter controlling the overall scale of the covariance.

Correlation parameter across categories, in the interval $(-1, 1)$. Used when comp_like = 3 or 4.

Correlation parameter across sexes, in the interval $(-1, 1)$. Used when comp_like = 4.

An RTMB object created via 'ADFun'.

An 'adnuts' object containing MCMC samples.

Character vector specifying the names of components in the model report to extract.

Number of cores to use

List of data from 'SPoRC'

List of n_models of 'SPoRC' report lists

Vector of model names

Array of observed compositions, dimensioned by [region, year, bin, sex, fleet]. May contain NAs, which are removed when filtering by years.

Array of expected compositions, dimensioned the same as obs_mat. May contain NAs, which are removed when filtering by years.

Input (or effective if Multinomial) sample size. Dimensions depend on comp_type:

• comp_type = 0 (aggregated): vector of length n_years.

• comp_type = 1 (split by region and sex): array [n_regions, n_years, n_sexes].

• comp_type = 2 (split by region, joint by sex): matrix [n_regions, n_years].

Dirichlet-multinomial overdispersion parameter(s). Dimensions must match N:

• aggregated: scalar

• split by sex: matrix [n_regions, n_sexes]

• joint by sex: vector of length n_regions

Logistic-normal covariance matrix. Dimensions depend on comp_type:

• aggregated: matrix [n_bins, n_bins]

• split by region and sex: array [n_regions, n_bins, n_bins, n_sexes]

• joint by sex: array [n_regions, n_bins, n_bins]

Use [get_logistN_Sigma()] to help construct this input.

Vector of years to filter to. Must match dimensions of obs_mat and exp_mat.

Fleet identifier (character or numeric) to filter to.

Vector of age or length bin labels corresponding to the composition categories.

Integer specifying how compositions are structured:

• 0 = aggregated across regions and sexes

• 1 = split by region and sex

• 2 = split by region, joint by sex

Character label describing whether bins represent ages or lengths.

Integer specifying the likelihood type (defaults to 0):

• 0 = multinomial

• 1 = Dirichlet-multinomial

• 2–4 = logistic-normal variants

Parameter list from setting up TMB object

Mapping list from setting up TMB object

SD Report from TMB obj

List. Data object containing ages, years, weight-at-age, maturity, natural mortality, and other simulation/assessment info.

List. Report object from RTMB containing estimated parameters like Fmort, selectivity, recruitment, steepness.

Numeric. Target Spawning Potential Ratio fraction. Required for SPR-based reference points.

Numeric. Mortality time until spawning.

Numeric vector. Female sex ratio by region.

Integer. First year used to compute mean recruitment.

Integer. Age at recruitment.

Character. "single_region" or "multi_region".

Character. Type of reference point:

SPR

Single-region SPR reference point

independent_SPR

Multi-region SPR without movement

global_SPR

Multi-region SPR with movement

BH_MSY

Single-region Beverton–Holt MSY

independent_BH_MSY

Multi-region BH-MSY without movement

global_BH_MSY

Multi-region global BH-MSY with movement

local_BH_MSY

Multi-region local BH-MSY with movement

Integer. Number of years to average demographic rates when calculating reference points.

Number of newton steps to take to solve for equilibrium recruitment in the origin region when local_BH_MSY is assumed.

Dataframe generated from do_retrospective

Age in which recruitment occurs

Dataframe outputted from do_retrospective function

List of n_models of 'SPoRC' report lists

Vector of model names

Character vector specifying whether to output age or length-based selectivity (age, length)

Year index for which selectivity year to plot (can be an integer or vector)

List of n_models of 'SPoRC' report lists

List of n_models of 'SPoRC' sdreport lists

Vector of model names

Boolean for whether confidence intervals are plotted

A list containing optimization results. Must include either:

• '"par"' and '"objective"' (e.g., from 'optim'), or

• '"par"' and '"value"' (e.g., from 'nlminb')

Numeric. Penalty multiplier for the number of parameters. Default is 2.

Numeric. Sample size. Default is 'Inf'.

List of n_models of 'SPoRC' data lists

List of n_models of 'SPoRC' report lists

List of n_models of sd report lists from 'SPoRC'

Character vector of model names

Path to the output directory. Users only need to specify the path.

List object obtained from get_osa, that contains a dataframe of residuals and aggregated fits.

sd report list from a 'SPoRC' model

report list from a 'SPoRC' model

Value for maximum gradient tolerance to use

Value for maximum standard error tolerance to use

Value for maximum correlation tolerance to use

Numeric value to constrain

Year index

Simulation index

Simulation environment will all the necessary elements to run the annual cycle

A list of model input data, including at least observed compositions ('ObsFishAgeComps', 'ObsFishLenComps', 'ObsSrvAgeComps', 'ObsSrvLenComps') and corresponding weights ('Wt_FishAgeComps', 'Wt_FishLenComps', 'Wt_SrvAgeComps', 'Wt_SrvLenComps').

A list of model parameters to be passed to [fit_model()].

A list or mapping object used to specify fixed or estimated parameters in [fit_model()].

A character string of random effects passed to [fit_model()].

Integer. Number of Francis reweighting iterations to perform. Default is '10'.

Integer. Number of Newton loops passed to [fit_model()]. Default is '0'.

Data list for RTMB model

Integer vector specifying which years to mark as unused. Only years present in data$years are considered.

Character vector specifying which data types to modify. Possible values include:

"Catch"

Catch data indicators.

"FishIdx"

Fishery index data indicators.

"FishAgeComps"

Fishery age composition data indicators.

"FishLenComps"

Fishery length composition data indicators.

"SrvIdx"

Survey index data indicators.

"SrvAgeComps"

Survey age composition data indicators.

"SrvLenComps"

Survey length composition data indicators.

"Tagging"

Tagging data and associated cohorts.

List containing data, parameter, and map lists for the model.

Numeric array of weight-at-age (spawning), dimensioned [n_regions, n_years, n_ages, n_sexes].

Numeric array of weight-at-age (fishery), dimensioned [n_regions, n_years, n_ages, n_sexes, n_fish_fleets].

Numeric array of weight-at-age (survey), dimensioned [n_regions, n_years, n_ages, n_sexes, n_srv_fleets].

Numeric array of maturity-at-age, dimensioned [n_regions, n_years, n_ages, n_sexes].

Numeric value for a constant to add to composition data. Default is 1e-3. Not used if logistic normal likelihoods are utilized.

Numeric value for a constant to add to composition data. Default is 1e-4.

Numeric value for a constant to add to composition data. Default is 1e-4.

Numeric value for a constant to add to composition data. Default is 1e-10

Numeric matrix or array representing the ageing error transition matrix. If a matrix (2D), dimensions should be [number of modeled ages, number of observed composition ages] and the ageing error is assumed to be constant over time. If an array (3D), dimensions should be [number of years, number of modeled ages, number of observed composition ages] allowing ageing error to vary by year. Defaults to an identity matrix (no ageing error) if not specified, assuming observed age bins exactly match modeled age bins.

**Note:** If the observed age composition bins differ from the modeled age bins (e.g., observed ages 2–10 while modeled ages are 1–10), the default identity matrix will cause a dimensional mismatch and misalignment. In such cases, users should provide a custom ageing error matrix mapping modeled to observed ages. For example, to drop the first modeled age bin, supply a matrix like diag(1, 10)[, 2:10]. This ensures proper alignment of age bins for likelihood calculations.

Integer flag indicating whether to apply a natural mortality prior (0 = no, 1 = yes).

Numeric vector of length two giving the mean (in normal space) and standard deviation of the natural mortality prior.

Integer flag indicating whether to fit length data (0 = no, 1 = yes).

Numeric array of size-at-age transition probabilities, dimensioned [n_regions, n_years, n_lens, n_ages, n_sexes].

Character string specifying whether selectivity is age or length-based. Default is age-based

• "length": Length-based selectivity.

• "age": Age-based selectivity

Character string specifying natural mortality estimation approach. Defaults to est_ln_M, which estimates mortality to be invariant, if blocks are not specified. Options:

• "est_ln_M": Estimates natural mortality across the defined natural mortality blocks.

• "fix": Fix all natural mortality parameters using the provided array.

Specification of age blocking for natural mortality estimation. Either a character string ("constant") or a list of index vectors, e.g., list(1:10, 11:30), which specifies 2 age blocks for M.

Specification of regional blocking for natural mortality. Either a character string ("constant") or a list of index vectors, e.g., list(1:3, 4:5), which specifies 2 region blocks for M.

Specification of year blocking for natural mortality. Either a character string ("constant") or a list of index vectors, e.g., list(1:10, 11:30), which specifies 2 year blocks for M.

Specification of sex blocking for natural mortality. Either a character string ("constant") or a list of index vectors, e.g., list(1:2), which specifies sex-invariant M.

Numeric array of fixed natural mortality values, dimensioned [n_regions, n_years, n_ages, n_sexes]. Required if M_spec = "fix".

Additional arguments for starting values such as ln_M and M_offset. These are ignored if M_spec = fix.

A list containing data, parameters, and map lists used by the model.

Numeric array of observed catches, dimensioned [n_regions, n_years, n_fish_fleets].

Catch units - Array dimensioned by n_regions x n_fish_fleets

• "abd": Catch units in abundance

• "biom": Catch units in biomass (default)

Integer matrix with dimensions [n_years, n_fish_fleets], specifying catch data types:

• 0: Use aggregated catch data for the year.

• 1: Use region-specific catch data for the year.

Indicator array [n_regions, n_years, n_fish_fleets] specifying whether to include catch data in the fit:

• 0: Do not use catch data.

• 1: Use catch data and fit.

Integer flag indicating whether to apply a fishing mortality penalty:

• 0: Do not apply penalty.

• 1: Apply penalty.

Integer flag indicating whether all regional fishing mortality deviations are estimated:

• 0: Some fishing mortality deviations are aggregated across regions.

• 1: All fishing mortality deviations are regional.

Character string specifying observation error structure for catch data. Default behavior fixes sigmaC at a starting value of 1e-3 (log-scale ln_sigmaC = log(1e-3)) for all regions, years, and fleets. Other options include:

• "est_shared_f": Estimate sigmaC shared across fishery fleets, unique by region and year.

• "est_shared_r": Estimate sigmaC shared across regions, unique by fleet and year.

• "est_shared_y": Estimate sigmaC shared across years, unique by region and fleet.

• "est_shared_r_f": Estimate sigmaC shared across regions and fleets, unique by year.

• "est_shared_f_y": Estimate sigmaC shared across fleets and years, unique by region.

• "est_shared_r_y": Estimate sigmaC shared across regions and years, unique by fleet.

• "est_shared_r_y_f": Estimate single sigmaC shared across regions, years, and fleets.

• "fix": Fix sigmaC at the starting value.

• "est_all": Estimate separate sigmaC for each region, year, and fleet combination.

Character string specifying process error structure for aggregated catch observation error. Default fixes sigmaC_agg at the starting value (log-scale sigmaC_agg). Other options include:

• "est_shared_f": Estimate sigmaC_agg shared across fishery fleets, unique by year.

• "est_shared_y": Estimate sigmaC_agg shared across years, unique by fleet.

• "est_shared_y_f": Estimate single sigmaC_agg shared across years and fleets.

• "fix": Fix at the starting value.

• "est_all": Estimate separate parameters for each year and fleet combination.

Character string specifying process error structure for fishing mortality. Default fixes sigmaF at 1 on the log scale (i.e., ln_sigmaF = 0). Other options include:

• "est_shared_f": Estimate sigmaF shared across fishery fleets.

• "est_shared_r": Estimate sigmaF shared across regions but unique by fleet.

• "est_shared_r_f": Estimate sigmaF shared across regions and fleets.

• "fix": Fix sigmaF at the starting value.

• "est_all": Estimate separate sigmaF for each region and fleet.

Character string specifying process error structure for aggregated fishing mortality. Default fixes sigmaF_agg at the starting value (log-scale ln_sigmaF_agg). Other options include:

• "est_shared_f": Estimate sigmaF_agg shared across fishery fleets.

• "fix": Fix at the starting value.

• "est_all": Estimate separate parameters for each fishery fleet.

Additional arguments specifying starting values for ln_sigmaC, ln_sigmaF, and ln_sigmaF_agg.

Numeric vector of years.

Numeric vector of age classes.

Numeric vector of length bins; can be set to 1 if length data are not modeled.

Integer specifying the number of spatial regions.

Integer specifying the number of sexes.

Integer specifying the number of fishery fleets.

Integer specifying the number of survey fleets.

Number of projection years for deviation parameters (ln_RecDevs, logit_move_devs, ln_fishsel_devs, ln_srvsel_devs)

Logical flag indicating whether to print progress messages (default FALSE).

List containing a data list, parameter list, and map list

Observed fishery index data as a numeric array with dimensions [n_regions, n_years, n_fish_fleets].

Standard errors associated with ObsFishIdx, also dimensioned [n_regions, n_years, n_fish_fleets].

Character vector of length n_fish_fleets specifying the type of index data. Options are "abd" for abundance, "biom" for biomass, and "none" if no index is available.

Logical or binary indicator array ([n_regions, n_years, n_fish_fleets]) specifying whether to include a fishery index in the likelihood (1) or ignore it (0).

Observed fishery age composition data as a numeric array with dimensions [n_regions, n_years, n_ages, n_sexes, n_fish_fleets]. Values should reflect counts or proportions (not required to sum to 1, but should be on a comparable scale).

Indicator array ([n_regions, n_years, n_fish_fleets]) specifying whether to fit fishery age composition data (1) or ignore it (0).

Input sample size for age compositions, array dimensioned [n_regions, n_years, n_sexes, n_fish_fleets]. Required if observed age comps are normalized (i.e., sum to 1), to correctly scale the contribution to the likelihood.

Observed fishery length composition data as a numeric array with dimensions [n_regions, n_years, n_lens, n_sexes, n_fish_fleets]. Values should reflect counts or proportions.

Indicator array ([n_regions, n_years, n_fish_fleets]) specifying whether to fit fishery length composition data (1) or ignore it (0).

Same as ISS_FishAgeComps, but for length compositions.

Character vector of length n_fish_fleets specifying the likelihood type used for fishery age composition data. Options include "Multinomial", "Dirichlet-Multinomial", and "iid-Logistic-Normal". Use "none" to omit the likelihood.

Same as FishAgeComps_LikeType, but for fishery length composition data.

Character vector specifying how age compositions are structured by fleet and year range. Options include:

• "agg": Aggregated across regions and sexes.

• "spltRspltS": Split by region and by sex (compositions sum to 1 within region-sex group).

• "spltRjntS": Split by region but summed jointly across sexes.

• "none": No composition data used.

Format each element as "<type>_Year_<start>-<end>_Fleet_<fleet number>" (e.g., "agg_Year_1-10_Fleet_1").

Same as FishAgeComps_Type, but for length compositions.

Additional arguments specifying starting values for overdispersion parameters (e.g., ln_FishAge_theta, ln_FishLen_theta, ln_FishAge_theta_agg, ln_FishLen_theta_agg).

List containing a data list, parameter list, and map list

Character vector specifying the form of continuous time-varying selectivity for each fishery fleet. The vector must be length n_fish_fleets, and each element must follow the structure: "<time variation type>_Fleet_<fleet number>".

Valid time variation types include:

• "none": No continuous time variation (default)

• "iid": Independent and identically distributed deviations across years.

• "rw": Random walk in time.

• "3dmarg": 3D marginal time-varying selectivity.

• "3dcond": 3D conditional time-varying selectivity.

• "2dar1": Two-dimensional AR1 process.

For example:

• "iid_Fleet_1" applies an iid time-varying structure to Fleet 1.

• "none_Fleet_2" means no time variation is used for Fleet 2.

Character vector specifying the fishery selectivity blocks for each region and fleet.

Each element must follow one of the following structures:

• '"Block_<block number>_Year_<start>-<end>_Fleet_<fleet number>"'

• '"Block_<block number>_Year_<start>-terminal_Fleet_<fleet number>"'

• '"none_Fleet_<fleet number>"'

This argument defines how fishery selectivity varies over time for each fleet:

• "Block_..." entries specify discrete time blocks during which selectivity parameters are assumed constant.

• "none_..." entries indicate that selectivity is constant across all years for the specified fleet.

If time-block-based selectivity is specified for a fleet (via fish_sel_blocks), its corresponding continuous selectivity option (in cont_tv_fish_sel) must be set to "none_Fleet_<fleet number>". The two approaches—blocked and continuous time-varying selectivity—are mutually exclusive. The default for each fleet is "none_Fleet_x" (i.e., no selectivity blocks).

Character vector specifying the fishery selectivity functional form for each fleet, and optionally by time block.

Each element must follow one of the following structures:

• "<selectivity model>_Fleet_<fleet number>"

• "<selectivity model>_Block_<block number>_Fleet_<fleet number>"

The first form applies a single selectivity model across all years for the specified fleet. The second form allows the user to assign a distinct selectivity model to a specific time block, as defined in fish_sel_blocks.

Available selectivity model types include:

• "logist1" — Logistic function with parameters a50 and k.

• "logist2" — Logistic function with parameters a50 and a95.

• "gamma" — Dome-shaped gamma function with parameters amax and delta.

• "exponential" — Exponential function with a power parameter.

• "dbnrml" — Double-normal function with six parameters.

If multiple selectivity time blocks are specified for a fleet (using fish_sel_blocks), then the corresponding selectivity model for each block must be explicitly defined using the "<model>_Block_<block>_Fleet_<fleet>" format. If blocks are not defined for a fleet, use the "<model>_Fleet_<fleet number>" format only. For mathematical definitions and implementation details of each selectivity form, refer to the model equations vignette.

Integer (0 or 1). Flag to enable/disable fishery catchability priors. When set to 1, applies log-normal priors to fishery selectivity parameters as specified in fish_q_prior. When set to 0, no priors are applied.

Data frame containing prior specifications for fishery catchability parameters. Must include columns: region (region index), fleet (fleet index), block (time block index), mu (prior mean on natural scale), and sd (prior standard deviation on log scale). Each row specifies a log-normal prior N(log(mu), sd) for a given catchability parameter. Only parameters with rows in this data frame will have priors applied.

Character vector specifying fishery catchability (q) blocks for each fleet. Each element must follow the structure: "Block_<block number>_Year_<start>-<end>_Fleet_<fleet number>" or "none_Fleet_<fleet number>". Default is "none_Fleet_x".

This allows users to define time-varying catchability blocks independently of selectivity blocks. The blocks must be non-overlapping and sequential in time within each fleet.

For example:

• "Block_1_Year_1-35_Fleet_1" assigns block 1 to Fleet 1 for years 1–35.

• "Block_2_Year_36-56_Fleet_1" continues with block 2 for years 36–56.

• "Block_3_Year_57-terminal_Fleet_1" assigns block 3 from year 57 to the terminal year for Fleet 1.

• "none_Fleet_2" indicates no catchability blocks are used for Fleet 2.

Internally, these specifications are converted to a [n_regions, n_years, n_fish_fleets] array, where each block is mapped to the appropriate years and fleets.

Character string specifying how process error parameters for fishery selectivity are estimated across regions and sexes. This is only relevant if cont_tv_fish_sel is not set to "none"; otherwise, all process error parameters are treated as fixed.

Available options include:

• "est_all": Estimates separate process error parameters for each region and sex.

• "est_shared_r": Shares process error parameters across regions (sex-specific parameters are still estimated).

• "est_shared_s": Shares process error parameters across sexes (region-specific parameters are still estimated).

• "est_shared_r_s": Shares process error parameters across both regions and sexes, estimating a single set of parameters.

• "est_shared_f_x": Shares process error parameters with another fleet, where x is the fleet number to share with. This option forces multiple fleets to have identical process error variance and correlation structures for their time-varying selectivity. For example, "est_shared_f_2" means the current fleet will use the same process error parameters as fleet 2. The reference fleet (fleet x) must use one of the other sharing options and cannot itself be sharing with another fleet.

• "fix" or "none": Does not estimate process error parameters; all are treated as fixed.

Character string specifying the structure for estimating fixed-effect parameters of the fishery selectivity model (e.g., a50, k, amax). This controls whether selectivity parameters are estimated separately or shared across regions and sexes.

Available options include:

• "est_all": Estimates separate fixed-effect selectivity parameters for each region and sex.

• "est_shared_r": Shares parameters across regions (sex-specific parameters are still estimated).

• "est_shared_s": Shares parameters across sexes (region-specific parameters are still estimated).

• "est_shared_r_s": Shares parameters across both regions and sexes, estimating a single set of fixed-effect parameters.

• "est_shared_f_x": Shares fixed-effect selectivity parameters with another fleet, where x is the fleet number to share with. This option forces multiple fleets to have identical selectivity curves by using the same underlying parameters (e.g., same a50, k, amax values). For example, "est_shared_f_2" means the current fleet will use the same fixed-effect selectivity parameters as fleet 2. The reference fleet (fleet x) must use one of the other sharing options and cannot itself be sharing with another fleet.

• "fix": Fixes all selectivity parameters to their initial values (no estimation).

• "none": No selectivity parameters are estimated (equivalent to "fix").

Character string specifying the structure of fishery catchability (q) estimation across regions. This controls whether separate or shared parameters are used.

Available options include:

• "est_all": Estimates separate catchability parameters for each region.

• "est_shared_r": Estimates a single catchability parameter shared across all regions.

Character string specifying the structure of process error deviations in time-varying fishery selectivity dimensioned by the number of fishery fleets. This determines how deviations are estimated across regions and sexes.

Available options include:

• "est_all": Estimates a separate deviation time series for each region and sex.

• "est_shared_r": Shares deviations across regions (sex-specific deviations are still estimated).

• "est_shared_s": Shares deviations across sexes (region-specific deviations are still estimated).

• "est_shared_r_s": Shares deviations across both regions and sexes, estimating a single deviation time series.

• "est_shared_f_x": Shares deviations with another fleet, where x is the fleet number to share with. This option allows multiple fleets to use identical deviation parameters, reducing the number of parameters to estimate. For example, "est_shared_f_2" means the current fleet will use the same deviation parameters as fleet 2. The reference fleet (fleet x) must use one of the other sharing options ("est_all", "est_shared_r", "est_shared_s", or "est_shared_r_s") and cannot itself be sharing with another fleet.

• "fix": Fixes all deviation parameters to zero (no time-variation).

• "none": No deviation parameters are estimated (equivalent to "fix").

This argument is only used when a continuous time-varying selectivity form is specified (e.g., via cont_tv_fish_sel).

Character string specifying which correlation structures to suppress when using semi-parametric time-varying selectivity models. Only used if cont_tv_sel is set to one of "3dmarg", "3dcond", or "2dar1".

This option allows users to turn off estimation of specific correlation components in the time-varying selectivity model. This can improve stability or enforce assumptions about independence in the temporal or age structure.

Available options:

• "corr_zero_y": Sets year (temporal) correlations to 0.

• "corr_zero_b": Sets age correlations to 0.

• "corr_zero_y_b": Sets both year and bin correlations to 0.

• "corr_zero_c": Sets cohort correlations to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond".

• "corr_zero_y_c": Sets year and cohort correlations to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond".

• "corr_zero_b_c": Sets bin (age) and cohort correlations to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond".

• "corr_zero_y_b_c": Sets all correlations (year, bin (age), and cohort) to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond"; equivalent to an iid structure.

These correlation-suppression flags are ignored when cont_tv_sel is set to any other value.

Integer (0 or 1). Flag to enable/disable fishery selectivity priors. When set to 1, applies log-normal priors to fishery selectivity parameters as specified in fish_selex_prior. When set to 0, no priors are applied.

Data frame containing prior specifications for fishery selectivity parameters. Must include columns: region (region index), fleet (fleet index), block (time block index), sex (sex index), par (parameter index), mu (prior mean on natural scale), and sd (prior standard deviation on log scale). Each row specifies a log-normal prior N(log(mu), sd) for one selectivity parameter. Only parameters with rows in this data frame will have priors applied.

Whether or not continuous fishery time varying selectivity penalties are applied (if cont_tv_fish_sel > 0)

Additional arguments specifying starting values for fishery selectivity and catchability parameters (fishsel_pe_pars, ln_fishsel_devs, ln_fish_fixed_sel_pars, ln_fish_q)

List containing data, parameter, and map lists for the model.

Integer flag (0 or 1) indicating whether recruits move. Default is 0 (do not move).

Integer flag (0 or 1) indicating whether to use a fixed movement matrix (1) or estimate movement parameters (0). Default is 0.

Numeric array for fixed movement matrix dimensioned by [n_regions, n_regions, n_years, n_ages, n_sexes]. Default is an array of ones.

Integer flag (0 or 1) indicating whether to use movement priors. Default is 0 (no priors).

Numeric vector or array specifying prior values for movement parameters. If a vector, a constant prior is applied across all dimensions.

Either:

• Character string "constant" for age-invariant movement (default), or

• A list of numeric vectors specifying age blocks sharing parameters.

For example, list(c(1:6), c(7:10), c(11:n_ages)) defines three age blocks where:

• ages 1 to 6 share parameters,

• ages 7 to 10 share parameters,

• ages 11 to n_ages share parameters.

To specify age-invariant movement, use either "constant" or list(c(1:n_ages)).

Either:

• Character string "constant" for time-invariant movement (default), or

• A list of numeric vectors specifying year blocks sharing movement parameters.

Either:

• Character string "constant" for sex-invariant movement (default), or

• A list of numeric vectors specifying sex blocks sharing movement parameters.

Character string specifying continuous varying movement type. Available options:

• "none"

• "iid_y" (iid deviations by year)

• "iid_a" (iid deviations by age)

• "iid_y_a" (iid deviations by year and age)

• "iid_y_s" (iid deviations by year and sex)

• "iid_a_s" (iid deviations by age and sex)

• "iid_y_a_s" (iid deviations by year, age, and sex)

Default is "none".

Character string specifying process error parameter sharing. Available options:

• "est_shared_r"

• "est_shared_a"

• "est_shared_s"

• "est_shared_r_a"

• "est_shared_a_s"

• "est_shared_r_s"

• "est_shared_r_a_s"

• "est_all"

• "fix"

• "none"

Default is "none".

Additional parameters such as starting values for move_pars, logit_move_devs, and move_pe_pars.

List containing data, parameters, and map lists used by the model.

Character string specifying the recruitment model. Options are:

• "mean_rec": Recruitment is a fixed mean value.

• "bh_rec": Beverton-Holt recruitment with steepness parameter.

Character string specifying recruitment density dependence, options: "local", "global", or NULL.

Integer specifying the recruitment lag duration relative to spawning stock biomass (SSB).

Integer flag (0 or 1) indicating whether to apply a prior on steepness h.

Data frame specifying beta prior distributions for the 'h_trans' parameters. Must include the following columns: - 'region': Integer region index corresponding to the element in 'h_trans' being penalized. - 'mu': Mean of the prior in normal space (used to calculate the corresponding beta distribution). - 'sd': Standard deviation of the prior in normal space. For each row, a beta distribution is scaled to the interval [0.2, 1], and the corresponding element of 'h_trans' is transformed to that scale and penalized using the log-density from the beta distribution.

Integer flag (0 or 1) indicating whether to apply a prior on recruitment proportions.

Scalar or array specifying prior values for recruitment proportion parameters. If scalar, a constant uniform prior is applied across all dimensions.

Integer flag (0 or 1) indicating whether to apply a recruitment bias correction ramp.

Numeric vector of length 4 defining the recruitment bias ramp periods:

• Element 1: End year of no bias correction period.

• Element 2: End year of ascending bias ramp period.

• Element 3: End year of full bias correction period.

• Element 4: Start year of final no bias correction period.

For example, with 65 years total, c(21, 31, 60, 64) means:

• Years 1–21: No bias correction.

• Years 22–31: Ascending bias correction.

• Years 32–60: Full bias correction.

• Years 61–63: Descending bias ramp.

• Years 64–65: No bias correction.

Numeric specifying the maximum bias correction to apply to the recruitment bias ramp (should be between 0 and 1)

Integer year indicating when sigmaR switches from early to late values (0 disables switching).

Integer specifying how many of the most recent recruitment deviations to not estimate. Default is 0.

Integer flag specifying initialization of initial age structure:

• 0: Initialize by iteration.

• 1: Initialize using a scalar geometric series (does not account for movement).

• 2: Initialize using a matrix geometric series (accounts for movement; default).

Integer flag specifying how initial age structure deviations should be initialized. Default is stochastic for all ages except the recruitment age and the plus group.

• 0: Equilibrium initial age structure.

• 1: Stochastic initial age structure for all ages, except for the plus group, which follows equilibrium calculations (geometric series)

• 2: Stochastic initial age structure for all ages

Numeric value specifying the initial fishing mortality proportion relative to mean fishing mortality for initializing age structure.

Character string specifying estimation of recruitment variability (sigmaR):

• NULL or "est_all": Estimate separate sigmaR for early and late periods.

• "est_shared": Estimate one sigmaR shared across periods.

• "fix": Fix both sigmaR values.

• "fix_early_est_late": Fix early sigmaR, estimate late sigmaR.

Character string specifying estimation of initial age deviations:

• NULL: Estimate deviations for all ages and regions.

• "est_shared_r": Estimate deviations shared across regions.

• "fix": Fix all deviations.

Character string specifying recruitment deviation estimation:

• NULL: Estimate deviations for all regions and years.

• "est_shared_r": Estimate deviations shared across regions (global recruitment deviations).

• "fix": Fix all recruitment deviations.

Character string specifying steepness estimation:

• NULL: Estimate steepness for all regions if rec_model == "bh_rec".

• "est_shared_r": Estimate steepness shared across regions.

• "fix": Fix steepness values.

If rec_model == "mean_rec", steepness is fixed.

Numeric fraction specifying spawning timing within the year.

Character string specifying sex ratio estimation scheme:

• "est_all" estimates sex ratio for all blocks and regions independently

• "est_shared_r" estimates sex ratio shared across regions but varying by block

• "fix" fixes all sex ratio (no estimation)

Character vector specifying blocks of years and regions for sex ratio. Format examples:

• "Block_1_Year_1-15_Region_1"

• "Block_2_Year_16-terminal_Region_2"

• "none_Region_3" (means no block, constant for that region; this is the default option)

Additional arguments specifying starting values for recruitment parameters such as ln_global_R0, Rec_prop, h, ln_InitDevs, ln_RecDevs, and ln_sigmaR.

List containing a data list, parameter list, and map list

Observed survey index data as a numeric array with dimensions [n_regions, n_years, n_srv_fleets].

Standard errors associated with ObsSrvIdx, also dimensioned [n_regions, n_years, n_srv_fleets].

Logical or binary indicator array ([n_regions, n_years, n_srv_fleets]) specifying whether to include a survey index in the likelihood (1) or ignore it (0).

Character vector of length n_srv_fleets specifying the type of index data. Options are "abd" for abundance, "biom" for biomass, and "none" if no index is available.

Observed survey age composition data as a numeric array with dimensions [n_regions, n_years, n_ages, n_sexes, n_srv_fleets]. Values should reflect counts or proportions (not required to sum to 1, but should be on a comparable scale).

Indicator array ([n_regions, n_years, n_srv_fleets]) specifying whether to fit survey age composition data (1) or ignore it (0).

Observed survey length composition data as a numeric array with dimensions [n_regions, n_years, n_lens, n_sexes, n_srv_fleets]. Values should reflect counts or proportions.

Indicator array ([n_regions, n_years, n_srv_fleets]) specifying whether to fit survey length composition data (1) or ignore it (0).

Input sample size for age compositions, array dimensioned [n_regions, n_years, n_sexes, n_srv_fleets]. Required if observed age comps are normalized (i.e., sum to 1), to correctly scale the contribution to the likelihood.

Same as ISS_SrvAgeComps, but for length compositions.

Character vector of length n_srv_fleets specifying the likelihood type used for survey age composition data. Options include "Multinomial", "Dirichlet-Multinomial", and "iid-Logistic-Normal". Use "none" to omit the likelihood.

Same as SrvAgeComps_LikeType, but for survey length composition data.

Character vector specifying how age compositions are structured by fleet and year range. Options include:

• "agg": Aggregated across regions and sexes.

• "spltRspltS": Split by region and by sex (compositions sum to 1 within region-sex group).

• "spltRjntS": Split by region but summed jointly across sexes.

• "none": No composition data used.

Format each element as "<type>_Year_<start>-<end>_Fleet_<fleet number>" (e.g., "agg_Year_1-10_Fleet_1").

Same as SrvAgeComps_Type, but for length compositions.

Additional arguments specifying starting values for overdispersion parameters (e.g., ln_SrvAge_theta, ln_SrvLen_theta, ln_SrvAge_theta_agg, ln_SrvLen_theta_agg).

List containing a data list, parameter list, and map list

Character vector specifying the form of continuous time-varying selectivity for each survey fleet. The vector must be length n_srv_fleets, and each element must follow the structure: "<time variation type>_Fleet_<fleet number>".

Valid time variation types include:

• "none": No continuous time variation. (default)

• "iid": Independent and identically distributed deviations across years.

• "rw": Random walk in time.

• "3dmarg": 3D marginal time-varying selectivity.

• "3dcond": 3D conditional time-varying selectivity.

• "2dar1": Two-dimensional AR1 process.

For example:

• "iid_Fleet_1" applies an iid time-varying structure to Fleet 1.

• "none_Fleet_2" means no time variation is used for Fleet 2.

Character vector specifying the survey selectivity blocks for each region and fleet.

Each element must follow one of the following structures:

• '"Block_<block number>_Year_<start>-<end>_Fleet_<fleet number>"'

• '"Block_<block number>_Year_<start>-terminal_Fleet_<fleet number>"'

• '"none_Fleet_<fleet number>"'

This argument defines how survey selectivity varies over time for each fleet:

• "Block_..." entries specify discrete time blocks during which selectivity parameters are assumed constant.

• "none_..." entries indicate that selectivity is constant across all years for the specified fleet.

If time-block-based selectivity is specified for a fleet (via srv_sel_blocks), its corresponding continuous selectivity option (in cont_tv_srv_sel) must be set to "none_Fleet_<fleet number>". The two approaches—blocked and continuous time-varying selectivity—are mutually exclusive. The default for each fleet is "none_Fleet_x" (i.e., no selectivity blocks).

Character vector specifying the survey selectivity functional form for each fleet, and optionally by time block.

Each element must follow one of the following structures:

• "<selectivity model>_Fleet_<fleet number>"

• "<selectivity model>_Block_<block number>_Fleet_<fleet number>"

The first form applies a single selectivity model across all years for the specified fleet. The second form allows the user to assign a distinct selectivity model to a specific time block, as defined in srv_sel_blocks.

Available selectivity model types include:

• "logist1" — Logistic function with parameters a50 and k.

• "logist2" — Logistic function with parameters a50 and a95.

• "gamma" — Dome-shaped gamma function with parameters amax and delta.

• "exponential" — Exponential function with a power parameter.

• "dbnrml" — Double-normal function with six parameters.

If multiple selectivity time blocks are specified for a fleet (using srv_sel_blocks), then the corresponding selectivity model for each block must be explicitly defined using the "<model>_Block_<block>_Fleet_<fleet>" format. If blocks are not defined for a fleet, use the "<model>_Fleet_<fleet number>" format only. For mathematical definitions and implementation details of each selectivity form, refer to the model equations vignette.

Integer (0 or 1). Flag to enable/disable survey catchability priors. When set to 1, applies log-normal priors to survey selectivity parameters as specified in srv_q_prior. When set to 0, no priors are applied.

Data frame containing prior specifications for survey catchability parameters. Must include columns: region (region index), fleet (fleet index), block (time block index), mu (prior mean on natural scale), and sd (prior standard deviation on log scale). Each row specifies a log-normal prior N(log(mu), sd) for a given catchability parameter. Only parameters with rows in this data frame will have priors applied.

Character vector specifying survey catchability (q) blocks for each fleet. Each element must follow the structure: "Block_<block number>_Year_<start>-<end>_Fleet_<fleet number>" or "none_Fleet_<fleet number>".

This allows users to define time-varying catchability blocks independently of selectivity blocks. The blocks must be non-overlapping and sequential in time within each fleet.

For example:

• "Block_1_Year_1-35_Fleet_1" assigns block 1 to Fleet 1 for years 1–35.

• "Block_2_Year_36-56_Fleet_1" continues with block 2 for years 36–56.

• "Block_3_Year_57-terminal_Fleet_1" assigns block 3 from year 57 to the terminal year for Fleet 1.

• "none_Fleet_2" indicates no catchability blocks are used for Fleet 2.

Internally, these specifications are converted to a [n_regions, n_years, n_srv_fleets] array, where each block is mapped to the appropriate years and fleets.

Character string specifying how process error parameters for survey selectivity are estimated across regions and sexes. This is only relevant if cont_tv_srv_sel is not set to "none"; otherwise, all process error parameters are treated as fixed.

Available options include:

• "est_all": Estimates separate process error parameters for each region and sex.

• "est_shared_r": Shares process error parameters across regions (sex-specific parameters are still estimated).

• "est_shared_s": Shares process error parameters across sexes (region-specific parameters are still estimated).

• "est_shared_r_s": Shares process error parameters across both regions and sexes, estimating a single set of parameters.

• "est_shared_f_x": Shares process error parameters with another fleet, where x is the fleet number to share with. This option forces multiple fleets to have identical process error variance and correlation structures for their time-varying selectivity. For example, "est_shared_f_2" means the current fleet will use the same process error parameters as fleet 2. The reference fleet (fleet x) must use one of the other sharing options and cannot itself be sharing with another fleet.

• "fix" or "none": Does not estimate process error parameters; all are treated as fixed.

Character string specifying the structure for estimating fixed-effect parameters of the survey selectivity model (e.g., a50, k, amax). This controls whether selectivity parameters are estimated separately or shared across regions and sexes.

Available options include:

• "est_all": Estimates separate fixed-effect selectivity parameters for each region and sex.

• "est_shared_r": Shares parameters across regions (sex-specific parameters are still estimated).

• "est_shared_s": Shares parameters across sexes (region-specific parameters are still estimated).

• "est_shared_r_s": Shares parameters across both regions and sexes, estimating a single set of fixed-effect parameters.

• "est_shared_f_x": Shares fixed-effect selectivity parameters with another fleet, where x is the fleet number to share with. This option forces multiple fleets to have identical selectivity curves by using the same underlying parameters (e.g., same a50, k, amax values). For example, "est_shared_f_2" means the current fleet will use the same fixed-effect selectivity parameters as fleet 2. The reference fleet (fleet x) must use one of the other sharing options and cannot itself be sharing with another fleet.

• "fix": Fixes all selectivity parameters to their initial values (no estimation).

• "none": No selectivity parameters are estimated (equivalent to "fix").

Character string specifying the structure of survey catchability (q) estimation across regions. This controls whether separate or shared parameters are used.

Available options include:

• "est_all": Estimates separate catchability parameters for each region.

• "est_shared_r": Estimates a single catchability parameter shared across all regions.

Character string specifying the structure of process error deviations in time-varying survey selectivity dimensioned by the number of survey fleets. This determines how deviations are estimated across regions and sexes.

Available options include:

• "est_all": Estimates a separate deviation time series for each region and sex.

• "est_shared_r": Shares deviations across regions (sex-specific deviations are still estimated).

• "est_shared_s": Shares deviations across sexes (region-specific deviations are still estimated).

• "est_shared_r_s": Shares deviations across both regions and sexes, estimating a single deviation time series.

• "est_shared_f_x": Shares deviations with another fleet, where x is the fleet number to share with. This option allows multiple fleets to use identical deviation parameters, reducing the number of parameters to estimate. For example, "est_shared_f_2" means the current fleet will use the same deviation parameters as fleet 2. The reference fleet (fleet x) must use one of the other sharing options ("est_all", "est_shared_r", "est_shared_s", or "est_shared_r_s") and cannot itself be sharing with another fleet.

• "fix": Fixes all deviation parameters to zero (no time-variation).

• "none": No deviation parameters are estimated (equivalent to "fix").

This argument is only used when a continuous time-varying selectivity form is specified (e.g., via cont_tv_srv_sel).

Character string specifying which correlation structures to suppress when using semi-parametric time-varying selectivity models. Only used if cont_tv_sel is set to one of "3dmarg", "3dcond", or "2dar1".

This option allows users to turn off estimation of specific correlation components in the time-varying selectivity model. This can improve stability or enforce assumptions about independence in the temporal or age structure.

Available options:

• "corr_zero_y": Sets year (temporal) correlations to 0.

• "corr_zero_b": Sets bin correlations to 0.

• "corr_zero_y_b": Sets both year and bin correlations to 0.

• "corr_zero_c": Sets cohort correlations to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond".

• "corr_zero_y_c": Sets year and cohort correlations to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond".

• "corr_zero_b_c": Sets bin (age) and cohort correlations to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond".

• "corr_zero_y_b_c": Sets all correlations (year, bin (age), and cohort) to 0. Only valid for cont_tv_sel = "3dmarg" or "3dcond"; equivalent to an iid structure.

These correlation-suppression flags are ignored when cont_tv_sel is set to any other value.

A named list of formulas specifying environmental covariate relationships for each region and survey fleet. Each element should be named using the convention '"Region_<region>_Fleet_<fleet>"' and contain a formula object using covariate names present in 'srv_q_cov_dat'. The formula determines how environmental covariates influence survey catchability. If 'NULL', no environmental covariate effects are included.

A named list containing time series vectors (typically by year) of environmental covariates used in the 'srv_q_formula'. Each entry should be a numeric vector of length equal to the number of years, and names must match the variable names used in the formulas. If 'NULL', survey catchability is assumed to be time-invariant (i.e., not influenced by environmental variables).

Integer (0 or 1). Flag to enable/disable survey selectivity priors. When set to 1, applies log-normal priors to survey selectivity parameters as specified in srv_selex_prior. When set to 0, no priors are applied.

Data frame containing prior specifications for survey selectivity parameters. Must include columns: region (region index), fleet (fleet index), block (time block index), sex (sex index), par (parameter index), mu (prior mean on natural scale), and sd (prior standard deviation on log scale). Each row specifies a log-normal prior N(log(mu), sd) for one selectivity parameter. Only parameters with rows in this data frame will have priors applied.

If both 'srv_q_formula' and 'srv_q_cov_dat' are non-'NULL', the model constructs time-varying design matrices for each region and fleet based on the provided formulas and environmental covariates. A coefficient array ('srv_q_coeff') and a mapping array ('map_srv_q_coeff') are created to estimate and track the associated regression coefficients. The design matrix is stored in 'srv_q_env', a 4D array indexed by [region, year, fleet, covariate].

If either argument is 'NULL', environmental covariate effects are excluded and survey catchability is treated as constant over time.

Important: All covariate time series in 'srv_q_cov_dat' must:

• Be numeric vectors with a length equal to the number of years in the model.

• Align to the same years across all covariates.

• Contain no missing values; users must impute or interpolate missing covariate values prior to use. For years in which the index is not used, values can be set at 0.

Covariates that are defined but not used in any formula can be filled with zeros (e.g., rep(0, n_yrs)). This avoids issues with list structure but does not affect the design matrix or model results.

Example formulas:

• "Region_1_Fleet_1" = ~ 0 + poly(env1_r1_f1, 3) + env2_r1_f1 uses a 3rd-degree polynomial for env1_r1_f1 and a linear term for env2_r1_f1.

• "Region_2_Fleet_1" = ~ 0 + env1_r2_f1 + env2_r2_f1 includes additive effects of two covariates.

• "Region_3_Fleet_2" = ~ NULL disables environmental covariates for that fleet-region.

Survey timing in fractions (n_regions * n_srv_fleets; default is 0.5)

Whether or not to apply continuous time-varying selectivity penalties (if cont_tv_srv_sel > 0)

Additional arguments specifying starting values for survey selectivity and catchability parameters (srvsel_pe_pars, ln_srvsel_devs, ln_srv_fixed_sel_pars, ln_srv_q, srv_q_coeff)

List containing a data list, parameter list, and map list

Numeric (0 or 1) indicating whether to use tagging data (1) or not (0)

Matrix [n_tag_cohorts x 2], where columns are release region and release year

Maximum number of years to track a tagged cohort

Array [n_tag_cohorts x n_ages x n_sexes] describing tagged fish releases

Array [max_tag_liberty x n_tag_cohorts x n_regions x n_ages x n_sexes] observed tag recaptures

Character string specifying tag likelihood type. One of:

• "Poisson"

• "NegBin"

• "Multinomial_Release"

• "Multinomial_Recapture"

• "Dirichlet-Multinomial_Release"

• "Dirichlet-Multinomial_Recapture"

Example: Tag_LikeType = "NegBin"

Numeric indicating minimum years post-release to include in fitting

Fractional year when tagging occurs (e.g., 0.5 for mid-year)

Character string specifying tag recovery selectivity. One of:

• "Uniform_DomFleet"

• "SexAgg_DomFleet"

• "SexSp_DomFleet"

• "Uniform_AllFleet"

• "SexAgg_AllFleet"

• "SexSp_AllFleet"

Example: tag_selex = "SexSp_AllFleet"

Character string specifying tag natural mortality parameterization. One of:

• "AgeAgg_SexAgg"

• "AgeSp_SexAgg"

• "AgeAgg_SexSp"

• "AgeSp_SexSp"

Example: tag_natmort = "AgeSp_SexSp"

Numeric (0 or 1) whether to use tag reporting rate prior

Data frame containing prior specifications for tag reporting parameters. Must include columns: region (region index), block (time block index), mu (Numeric mean for tag reporting prior (normal space); NA if symmetric beta is used), sd (Numeric standard deviation for tag reporting prior (normal space)), and type (0 == symmetric beta, 1 == regular beta). Each row specifies a beta prior for one tag reporting parameter. Only parameters with rows in this data frame will have priors applied.

List or character specifying pooling of tagging data by age groups. Examples:

• list(1:5, 6:11, 12:20) pools these age groups together

• "all" pools all ages together (internally converted to list(1:n_ages))

• as.list(1:n_ages) fits each sex separately