Project number: 2017-102
Project Status:
Budget expenditure: $188,995.00
Principal Investigator: Paul Burch
Organisation: CSIRO Oceans and Atmosphere Hobart
Project start/end date: 4 Feb 2018 - 29 Sep 2018


The Status of Australian Fish Stocks project is increasing the number of species/stocks to be included but many of these new stocks may fall into the "undefined" category and, because they suggest a lack of assessment and management, they lower the overall impression of the state of fisheries management within Australia. The FRDC National Priority 1 has two targets relating to the "undefined" category. By 2020, the target is to increase the number of species covered in SAFS to 200, and at the same time, to reduce the percentage of stocks classified as undefined to less than 10%. Most major commercial species by value are already included in SAFS, so increasing that number to 200 will mean including many data-poor fisheries making achieving both targets by 2020 difficult. An earlier project (2016-135) disarticulated the undefined category into sub-groups at least one of which should be amenable to data-poor assessment techniques. There is thus a need, within each jurisdiction, to identify which of the new species selected for inclusion in SAFS are likely to be classed as undefined and yet still amenable to a data-poor assessment method. There have been many recent developments with data-poor stock assessment methods and there is thus also a need in all jurisdictions for staff training to develop more local expertise in these new methods and to transfer suitable custom software for conducting such analyses. With the agreement of all parties involved there is a need to then apply and document the particular assessment method used that permits a status determination for each species selected, thereby reducing the number of undefined species.


1. Of the species proposed for inclusion in the 2018 and 2020 SAFS reports, identify those which may be deemed 'undefined' in each jurisdiction and yet potentially amenable to a data-poor stock assessment.
2. In each jurisdiction with potentially ‘undefined’ species, arrange a training workshop for local staff using the candidate species from objective 1 to act as case studies for the application of suitable data-poor stock assessment methods.
3. Include the 15 potentially assessable species from SAFS 2016, as identified in Phase one of this project (FRDC Project 2016-135).
4. Ensure that at least the local scientists involved with SAFS assessments understand how to use the illustrated data-poor assessment methods to develop a defensible stock status report and, if required, associated management advice.

Final report

ISBN: 978-1-4863-1288-7
Authors: Malcolm Haddon Paul Burch Natalie Dowling Rich Little
Final Report • 2019-07-01 • 1.93 MB


Seven data-poor assessment method training workshops were run in seven different jurisdictions (Tasmania, Victoria, South Australia, Western Australia, Northern Territory, Queensland, and New South Wales). Originally the workshops were to have been undertaken from March to the end of May 2018. However, the availability of project staff combined with the availability of people within the jurisdictions meant that time-table had to be extended into June.
Two open source R packages, simpleSA and cede, were used in the workshops, with additional development of the software contained in each one continuing as experience in the different jurisdictions expanded. cede contained software to assist with data exploration (simple mapping and data summary functions) and with illustrating and comparing different catch-effort standardization techniques. simpleSA contained three main data-poor stock assessment techniques (catch-MSY, surplus-production modelling, and age-structured surplus production modelling) plus functions to assist with catch-curve analysis.
The workshops consisted of an introduction to the problems of assessing data-poor fisheries, potential solutions, and their implications for management. The workshops included an introductory lecture and then live demonstrations of the software with expla-nations of the limitations and assumptions of each approach, followed by hands-on use by participants using either data sets included in the packages or, ideally, their own datasets prepared before the workshops.
Initially the workshops were designed around the idea of being two days long, but after the first two workshops, this was altered to become three days (for all but the Tasmanian workshop, which required only 2 days). This allowed time for participants to more fully explore their data, to make brief presentations of analyses they had conducted, and to receive feedback on these from the workshop presenters and their own colleagues.

Related research