Project number: 2005-063
Project Status:
Budget expenditure: $576,428.80
Principal Investigator: Norman G. Hall
Organisation: Murdoch University
Project start/end date: 29 Sep 2005 - 30 Aug 2010


There are currently no standard techniques that can be used to assess whether fishing has had significant impacts on ecosystem structure. The current round of EPBC assessments has demonstrated that a more robust assessment of ecosystem impacts will be required when the next application for export exemption is submitted in five years time. It is vital for WA's export fisheries that the types of changes in exploitation and/or environment that could cause marked changes in ecosystem structure are identified, the types of data necessary to assess whether such changes are occurring are determined, and cost-effective methods are developed to provide information on the level of ecosystem changes that have occurred.

Having appropriate techniques to assess whether fishing within a region is significantly modifying the ecosystem is seen as a national priority. At a recent workshop held by the Research Committee of the Australian Fisheries Managers Forum (AFMF) it was concluded that different types and levels of analysis are likely to be needed for different purposes and it would be “inappropriate to abandon any particular approach prematurely”. Given the potential costs associated with the collection of new or additional data at an ecosystem level, it is imperative that attempts are made to see if existing datasets, such as those currently maintained by fisheries agencies on catch and effort, are suitable for this purpose and to apply modelling approaches, such as those developed in FRDC 2000/311, to these refined datasets.

Completing routine ecosystem-level assessments will only be feasible when tools are available to simplify the complex process of analysing the multi-sector, multi-species databases that are present in WA (and most jurisdictions). There is a need, therefore, to identify and test a variety of statistical methods using these datasets to determine if any are suitable for detecting shifts in ecosystems or community structure.


1. Test the robustness of statistical procedures to identify impacts of multi-sector fishing on community composition using existing fishery data.
2. Assess the level of change in community composition in each bioregion of WA during the previous 30 years.
3. Identify key data to which ecosystem structure and management strategies are most sensitive and which should be collected in the future.
4. Identify critical changes in exploitation and/or environment that would impact marine ecosystems markedly.
5. Identify areas where more detailed research and/or monitoring are needed.

Final report

ISBN: 978-1-921845-04-8
Author: Norm Hall
Final Report • 2011-04-06 • 4.13 MB


Diversity and ecosystem-based indicators were calculated for commercial finfish fisheries from 1976 to 2005 for the West Coast, South Coast, Gascoyne, Pilbara and Kimberley bioregions. The ecosystem-based indices, which detect changes in the species composition of the food web within the ecosystem, were mean trophic level (1=herbivores to 5= peak predators), mean size of the fish in the catch (calculated using the maximum lengths for the species), and a Fishery-in-Balance (FIB) indicator.  The latter adjusts the magnitude of annual catch to account for changes in observed mean trophic level to determine whether the scaled catch has increased or decreased relative to that within a reference year.  The time series of ecosystem-based indices demonstrated that, in each bioregion, both the mean trophic level and the mean maximum length of catches have increased; possibly because the fisheries in some of these bioregions have developed and expanded spatially over the period from 1976 to 2005.  In the West and South Coast bioregions, the series appear to have stabilized, but they continue to increase in the other bioregions. There is no evidence from the commercial fishery data that, from 1976 to 2005, there has been any reduction in trophic level or mean maximum length that would be expected from fishing down the food web, and thus, it appears that, at this time, ecosystem services have not been affected by fishing or other factors.  It is possible that the indices are being maintained by continued spatial expansion of fishing and/or changes in targeting, and that, if exploitation increases and expansion is no longer possible, the ecosystem-based indices will stabilize and begin to decline.

Statistical analysis of the Western Australian (WA) data using the software package, Primer, demonstrated, however, that the species composition of the catches reported by fishers within each of the bioregions had changed over time. The species that most characterized the changes were identified.  The analysis was unable, however, to distinguish whether change in species composition and abundance resulted from fishery practice, recording and reporting processes, management changes, changes in exploitation or targeting, environmental change or a combination of these factors. Thus, while change in species composition had occurred in each bioregion, it was possible that this was due simply to expansion of fisheries to exploit different species groups in different locations within each bioregion. It is also possible that improved reporting by fishers, i.e. reporting of catches at species rather than family level, may also have contributed to the apparent change in species composition.

This and other studies have demonstrated that data collected for key fished and non-fished stocks within an ecosystem should include time series of total catches and reliable relative abundance indices, samples of age, length and sex composition representative both of the catches of each fishing sector and of the wild stocks, and data from studies of population biology, i.e. growth, maturity, sex change, reproduction.  Where appropriate and cost-effective, fishery-dependent data should be augmented by fishery-independent relative abundance, age composition and biological data.  Limited recreational catch data are currently available, and current estimates of abundance, i.e. cpue data from commercial fishers, are likely to be influenced greatly by changes in fishing power and targeting by fishers. 

Management strategies for the West Coast Bioregion were explored in this study. Results of this exploration demonstrated that the indicators, reference points and decision rules that have been adopted by the Department of Fisheries Western Australia for the demersal scalefish fisheries of the West Coast Bioregion are likely to be highly effective. Thus, for Western Australia’s finfish fisheries, fishing mortality estimates appear currently to be more reliable indicators of fishery status than abundance estimates, where the reference points for those indicators are those determined from the estimates of natural mortality for the different species.  Reference points for spawning biomass such as maximum sustainable yield and virgin spawning biomass rely to a much greater extent on trends in relative abundance, estimates of which are unreliable due to a paucity of accurate abundance data. 

Environmental change may affect the growth, reproduction, and carrying capacity of the various stocks. Changes in growth and reproduction can be monitored by appropriate data collection and analysis using traditional methods of fish population biology.  Changes in carrying capacity will be more difficult to assess as determination of a stock-recruitment relationship is typically difficult to determine, even when this is assumed to be constant. Although it was not possible to distinguish between fishery and environmental effects, the study demonstrated that the management strategies, which had been accepted for use in the demersal scalefish fishery of the West Coast Bioregion, would be likely to continue to be effective if the species were affected by changes in biological characteristics or carrying capacity.

Keywords: Ecosystem, trophic level, mean maximum length, species composition.

Related research