Stock assessments are the cornerstone of sustainable fisheries. However, these involve computationally intense methods to model fish population numbers, accounting for available fishery and biological information such as natural mortality, the age structure of the stock, age at spawning, fecundity and the ratio of males to females.
By Gio Braidotti
Over the years, a plethora of computer programs have been written to perform single-species stock assessments, but increasingly, off-the-shelf (OTS) software packages have become freely available worldwide.
Cathy Dichmont, an internationally recognised expert in stock assessment and natural resource management, says OTS packages offer considerable efficiencies, but there is a problem: “A review I undertook for the FRDC in 2016 found uptake of OTS software packages was low in Australia, particularly in comparison to other countries, such as the USA. For example, the review found that of the 76 data-rich assessment tools developed in Australia, about 52 of these could have been undertaken using OTS packages.”
The key barrier preventing uptake was a lack of knowledge about existing OTS packages, along with difficulty understanding where these packages are located within the World Wide Web.
“We provide information that can help decide if these packages are useful, we rate them in each model category defined on the website, and also point to key resources for each package.” – Cathy Dichmont
Those challenges have now been addressed with the 2020 launch of a website hosted by the FRDC called the Stock Assessment Toolbox. The site not only lists all available OTS packages within one website, but also reviews their functionality and curates their ongoing evolution.
It is the only site of its kind in the world that focuses on globally available packages for data-limited to data-rich stock assessment methods.
The development work was funded by the FRDC, CSIRO and Cathy Dichmont Consulting. It was undertaken by Cathy Dichmont and three CSIRO researchers:
- André Punt (Washington, USA), who is a professor at the University of Washington, where he develops new stock assessment methods;
- Natalie Dowling (Tasmania), who develops harvest strategies in data-limited contexts; and
- Roy Deng (Queensland), who specialises in fishery model programming, spatial analysis and fishery data management.
The project team conducted an in-depth search of freely available packages. More than 130 packages were initially identified. However, over 60 were removed from the list because they were no longer supported, had been replaced by a more modern package, or were still under development and not available for general use.
The Stock Assessment Toolbox.
On the positive side, the team observed that packages are increasingly being tailored for use in data-limited and data-moderate fisheries, creating more resources for these kinds of fisheries. This is a development that Cathy Dichmont strongly welcomes.
Prior to the launch, the site underwent independent testing by researchers in the fisheries sector. At launch, the site contained about 70 packages and three main resources:
- a package list and search function to help find an appropriate package for a fishery;
- simulated test data that can be used to test packages, given that the stock status of the simulated dataset is known; and
- installation and use guidelines for a set of key data-limited and data-moderate packages using the simulated test data.
“The Stock Assessment Toolbox provides detailed descriptions of what the listed packages do and what they are for,” Cathy Dichmont says.
“We provide information that can help decide if these packages are useful, we rate them in each model category defined on the website, and also point to key resources for each package.”
FRDC Research Codes: 2014-039, 2018-148
More information
Cathy Dichmont, cathydichmont@gmail.com
