In pbs-assess/PACea: An R package of Pacific ecosystem information to help facilitate an ecosystem approach to fisheries management

Introduction

## What is a stock assessment?

- Collect data from surveys and commercial fisheries, including ages of fish
- Data inform statistical population models to estimate historical biomass
- Models provide advice to fisheries managers

Pacific Hake assessment from 2023, estimate the historical population size:

plot(hake_biomass,
     y_max = 3.5)

## What is a stock assessment?

Advice often in the form of 'decision tables' that indicate probabilities of future events given
different catches:

| Catch (t) in 2023 | Probability spawning biomass declines from 2023 to 2024 | Probability spawning biomass in 2024 falls below 40% of unfished biomass |
| ----------------: | ------------------------------------------------------: | -----------------------------------------------------------------------: |
| 0 | 50% | 2% |
| 180,000 | 72% | 3% |
| 225,000 | 75% | 3% |
| 320,000 | 78% | 3% |
| 430,000 | 85% | 5% |

## Motivation

- Revised Fisheries Act: “... the Minister shall take into account the environmental conditions affecting a fish stock."

- `r colorize("Yet <50\\% of DFO’s stock assessments currently use environmental data.", "blue")`

- Only 28% of assessments in Pacific Region use environmental data.

- `r colorize("Leading cause of not using environmental data is availability of the data.", "red")`

  \

\footnotesize

Kulka DW, Thompson S, Cogliati K, Olmstead M, Austin D, Pepin D. (2022). An Accounting of Integration of Environmental Variables in Fishery Stock Assessments in Canada. Can. Tech. Rep. Fish. Aquat. Sci. 3473: viii + 79 p.
https://publications.gc.ca/collections/collection_2022/mpo-dfo/Fs97-6-3473-eng.pdf

\normalsize

##

knitr::include_graphics(paste0(here::here(),
                               "/talks/talks-manual-figures/motivation-1.png"))

##

knitr::include_graphics(paste0(here::here(),
                               "/talks/talks-manual-figures/motivation-2.png"))

## Motivation (based on a true story)

A search for sea surface temperature yields an overwhelming number (341) of choices:

knitr::include_graphics(paste0(here::here(),
                               "/talks/talks-manual-figures/motivation3.png"))

Likely requires extensive data wrangling to be usable, which usually takes way,
way, longer than anticipated.

So the SST analysis did not happen.

## Motivation

- Initial motivation was Dan Duplisea's `gslea` package for the Gulf of St. Lawrence.

- Outputs from a BC physical biogeochemical model have been shared by Angelica Peña.

- But the netCDF format may be unfamiliar to non-oceanographers --
  we take care of the wrangling into R.

- Open Data is great, but can be hard to
  `r colorize("convert raw data into usable information", "red")`.

- Primary audience is DFO stock assessment scientists, but usable by anyone
  (with a minimal working knowledge of R).

  \

\footnotesize

Duplisea et al. (2020). gslea: the Gulf of St Lawrence ecosystem
approach data matrix R-package. R package version 0.1 https://github.com/duplisea/gslea

\normalsize