mmsyFM

Multispecies MSY fleet model. Simulates mixed fisheries in which fleet segments allocate their fishing days across métiers by expected profit. It finds safeguarded multispecies MSY, and tunes and tests effort- and closure-based harvest strategies under uncertainty, with a Shiny app for participatory workshops. Documentation: https://henning-winker.github.io/mmsyFM/.

mmsyFM grew out of the logbook simulator of cpueTargeting and extends the shared-q model of the multispecies MSY app mmsyapp (see Authors and origins). In a shared-q model every effort rule moves along one line, because F_i / F_j = q_i / q_j is fixed. Here fishers choose where to fish.

Installation

# install.packages("remotes")
remotes::install_github("Henning-Winker/mmsyFM")

Quick start

library(mmsyFM)
sys <- atlantic_system()                                    # 10 stocks, 6 métiers, 3 segments
sq  <- ms_status_quo(sys, capacity = 1000, gamma = 1.5)      # open-access status quo
opt <- ms_optimise(sys, "yield", levers = "total", gamma = 1.5)  # MMSY with every stock >= Bpa
plot_kobe(list("Status quo" = sq, "Safeguarded MMSY" = opt), sys)

Articles:

• vignette("mmsyFM"): step by step through the functions, from the status quo to harvest strategies, performance statistics and plots;
• vignette("scenario-setup"): how the Atlantic and Gulf of Thailand systems were built and tuned, with a checklist for your own system;
• vignette("how-it-works"): the model, with equations and algorithms;
• vignette("management-options"): the management options in a real-world context.

Shiny app

mmsyFM::run_app()

The app’s tabs:

1. Set-up: the Atlantic example, the Gulf of Thailand draft or uploaded tables; capacity, profit preference, tuning of the daily costs.
2. Equilibrium: days by segment and open metiers.
3. Yield curves: shared q, profit-blind and profit-seeking fleets, closures.
4. Safeguarded MMSY: total yield, revenue, profit or the catch of one stock, with every stock at or above Bpa.
5. Harvest strategy: compare rules (no control, fixed MMSY effort, feedback rules, stock and metier closures) under uncertainty, with tables, projections, boxplots, trade-offs and decision tables.
6. Tune triggers.
7. Management options and Guide: the options in a real-world context, and the model with its equations.

Model

Layer	Content
Stocks	One Schaefer pool per stock; monthly catches; seasonal availability; JABBA-type process error, scaled by r (proc_sd_from_r()) and depletion (JABBA proc.dyn, relsig_jabba())
Métiers	Tactic × ground; catchability matrix by stock
Fleet	Segments with home métiers, loyalty, efficiency, daily cost and lognormal vessel efficiency; days follow the share of vessels that break even
Behaviour	P(métier) ∝ max(efficiency × expected revenue − cost, 0)^γ
Management	Effort caps by segment; métier closures (fractional, e.g. seasonal or area share); feedback rules (hcr_effort()) with assessment and implementation error, effort noise, a data lag, complexes, assessed subsets and nominal assignment
Tuning	ms_tune() (fleet composition at open access), ms_tune_hcr() (triggers to a risk limit, common random numbers)

Inputs

A system is four tables (see read_ms_system(); the Atlantic and Gulf of Thailand examples are in inst/extdata):

• species: species, code, group, r, K, price; optional season_amp, season_peak, F_cap (persistence floor, in Fmsy) and dep_init (starting B/K);
• metiers: metier, tactic, ground; optional cost;
• catchability: metier plus one column per stock, written directly or derived from a habitat landscape with ms_landscape();
• segments: segment, share_target, loyalty, efficiency, cost_multiplier, vessel_cv, home; optional capacity.

See vignette("scenario-setup") for a worked example (Gulf of Thailand) and vignette("how-it-works") for the model equations.

Roadmap

• Shiny app (run_app()): set-up, equilibrium, yield curves, safeguarded MMSY, harvest strategies with rule comparison, trigger tuning, guide.
• Gulf of Thailand draft (got_system_draft()): 20 stock groups; trawl, purse seine and small-scale fisheries with coastal zoning; conditioned on the 2024 depletions (ms_condition()).
• Monthly (seasonal) closures; habitat landscape for catchability (ms_landscape()).
• Performance statistics in the mseviz format (ms_perf()), boxplots, trade-offs and colour-coded decision tables (plot_quilt()).
• CPUE-based rules (observation model); SSCOM-type effort dynamics (Thorson et al. 2013); real prices and costs for the Gulf of Thailand.

Authors and origins

Authors: Henning Winker (maintainer), Ricardo Amoroso, Maite Pons and Rishi Sharma.

mmsyFM combines two lines of work, and we want to be transparent about where its parts come from.

1. The multispecies MSY app (mmsyapp). The starting point for the management questions is the multispecies MSY Shiny application by Ricardo Amoroso, Maite Pons and Rishi Sharma (mmsyapp). It represents a mixed fishery as Schaefer surplus-production stocks fished by one pool of effort with a fixed catchability per stock (shared q), and was applied to species groups of the Gulf of Thailand. mmsyFM keeps:

• the multispecies MSY framing: yield over effort with a safeguard for every stock;
• the shared-q model, as one of the treatments compared in the yield curves (ms_shared_q_curve());
• the Gulf of Thailand species groups and their r, K and 2024 depletions (got_species(), got_system_draft()).

2. The CPUE targeting simulator (cpueTargeting). Fleet behaviour comes from the multispecies logbook simulator of the cpueTargeting package (H. Winker), which re-implements the spatial simulator of Thorson et al. (2017), itself generalising Ono et al. (2015), with fisher targeting as in Winker et al. (2013, 2014). mmsyFM takes from it:

• métiers as fishing tactic x ground, and catchability from the overlap of where stocks live and where metiers fish (the Atlantic example was derived from its habitat landscape; ms_landscape() generalises this);
• targeting by expected profit, and vessel heterogeneity in efficiency (with a lower efficiency for generalist, mixed or small-scale vessels);
• the Schaefer stocks with seasonal availability that drive its catch rates.

New in mmsyFM: fleet segments with loyalty, costs and entry and exit; calibration of the fleet (ms_tune()) and of stock status (ms_condition()); the safeguarded MMSY optimiser; effort and closure rules in closed loop with assessment and implementation error, data lags, assessed subsets, complexes and nominal assignment; JABBA-type process error scaled by r and depletion; the F cap; and the performance statistics, whose presentation follows the FLR package mseviz (Mosqueira) and Winker et al. (2025).

References

Ono, K., Punt, A.E. and Hilborn, R. (2015) Think outside the grids: an objective approach to define spatial strata for catch and effort analysis. Fisheries Research 170, 89-101.

Thorson, J.T., Fonner, R., Haltuch, M.A., Ono, K. and Winker, H. (2017) Accounting for spatiotemporal variation and fisher targeting when estimating abundance from multispecies fishery data. Canadian Journal of Fisheries and Aquatic Sciences 74, 1794-1807.

Winker, H., Kerwath, S.E. and Attwood, C.G. (2013) Comparison of two approaches to standardize catch-per-unit-effort for targeting behaviour in a multispecies hand-line fishery. Fisheries Research 139, 118-131.

Winker, H., Kerwath, S.E. and Attwood, C.G. (2014) Proof of concept for a novel procedure to standardize multispecies catch and effort data. Fisheries Research 155, 149-159.

Winker, H., Cardinale, M., Gerritsen, H., Herrera, J.G., Castellanos, P., Farias, I., Sampedro, P., Moura, T., Mosqueira, I. and Kell, L. (2025) The ICES MSY approach to reference point estimation is not precautionary. ICES Journal of Marine Science 82, fsaf204.

Contact: Henning Winker, FAO Fisheries and Aquaculture Division (NFIFM), henning.winker@fao.org