ASFB Home > 2001 > Nearshore and estuarine fisheries – or 'How did we get in such a state'?

Nearshore and estuarine fisheries – or 'How did we get in such a state'?

Norm Hall

(Norm Hall is at the Centre for Fish & Fisheries Research, School of Biological Sciences and Biotechnology, Murdoch University.

He can be contacted at: normhall@central.murdoch.edu.au)

Abstract

The estuarine and nearshore fisheries of Western Australia may be characterised as multispecies, multigear, multisector finfish fisheries that are of relatively low value compared with the State’s major rock lobster, prawn and shellfish fisheries. These fisheries are the target of most of the recreational fishing within the State, and are thus more likely to be affected by any increase in recreational fishing effort. While commercial catch and effort data for these fisheries are available, such data have been difficult to obtain for the recreational sector. Furthermore, although biological data are available for many of the more important fish species in the fisheries, biological data for other species are limited as are data on age and length compositions of the stocks involved. Past stock assessments for these fisheries have often ignored the lack of data from the recreational fishery, or have made strong assumptions to overcome the deficiencies. However, the inadequacies of the data and resulting imprecision of stock assessments can no longer be overlooked. Strategies to improve data collection, and models that are capable of using the types of data that can be obtained from these fisheries, are urgently required.

History of Western Australian stock assessments for estuarine and nearshore fisheries

The estuarine and nearshore commercial and recreational fisheries of Western Australia are invariably multispecies and focus on a variety of finfish, which are fished with a number of different fishing gears. The recreational fishery is a significant component of these fisheries, taking a considerable proportion of the annual catch. The nearshore fish stocks and the fisheries that exploit these stocks have a wide spatial extent and their distribution is heterogeneous, with fishing activity often concentrated around human population centres. Although the fisheries are of considerable importance to the community, they have relatively low value compared with the major fisheries of Western Australia, such as those that exploit rock lobster, prawn, scallop and abalone.

As a consequence of their value, considerably less investment has been made in obtaining research information for these fisheries than has been applied in the major fisheries of the State. Catch and effort data from commercial fishers are often obtained for the estuarine and nearshore fisheries from a small number of fishers in mandatory monthly reports to the Department of Fisheries Western Australia. Because of the small number of fishers, the data are often imprecise and provide only sparse coverage of the fish stocks and, furthermore, confidentiality concerns restrict access and publication of the commercial fisheries data. Data on discarded or released fish are rarely obtained.

It is only in the last decade that a strategy has been implemented to obtain regular catch and effort data from recreational fishers. Creel census surveys that have been introduced by the Western Australian Department of Fisheries are intended to provide information on the annual fishing activities of recreational fishers at intervals of approximately five years. These surveys have been supplemented by the recently completed National Recreational Fishing Survey (Fisheries Research and Development Corporation project), which provides a snapshot of recreational fishing throughout Australia.

Limited biological data often exist for the species that are exploited in these estuarine and nearshore fisheries. While information on weight-length relationships, growth curves and reproduction exists for the more important of the fished species, data are often lacking for many of the less important. Monitoring of the length composition of commercial catches for these fisheries has occurred occasionally on an ad hoc basis yet, until the introduction of the regular creel census surveys of the recreational fishery, limited data on the length composition of recreational catches were collected. Age composition data on the species, if they exist at all, are limited. Details of the fishing gear used in the fisheries, its selectivity, and/or of the technology that has been introduced to improve the efficiency of fishing methods, are rarely recorded or are unavailable. However, such gaps in knowledge are being addressed through research projects on some key species identified by fisheries managers and fishers as requiring study, and which have been funded by the Fisheries Research and Development Corporation.

Clearly, such inadequate data pose challenges for stock assessments, which must inevitably yield imprecise estimates of the current status of each of the finfish stocks. Prior to the 1980s, single species methods were applied to these stocks, thus ignoring the complexity of interactions between species or technological interactions that are brought about by fishers changing their target species or area of operations. Occasionally, the data for several species were combined and treated as those for a single 'species'. With limited recreational data, stock assessment was often undertaken using only the data from the commercial fishing sector, thereby assuming that recreational effort and catch would have been proportional to the commercial effort and catch. To overcome the difficulties of different fishing gear, the equivalent effort that would have been required to take the catch obtained using a different fishing gear was calculated by dividing that catch by the catch rate of a standard gear, where the latter was usually the dominant fishing gear used by commercial fishers. However, the use of such effort measures usually ignored differences in the selectivities of the different fishing gears. Indeed, in many of the analyses, biomass dynamics models were applied, and gear selectivity was ignored under the assumption that it remained constant. The occasional measures of the recreational catch that became available were used to fix the magnitude of the recreational catch and allow extrapolation of the likely recreational catches in other years for which such data were not available.

More recently, in recognition of the multispecies, multisector nature of these estuarine and nearshore fisheries and the inadequacy of the data on the recreational catch and effort, time series approaches have been used in assessing the state of the stocks. These time series have attempted to use trends in catch rate and in catches to identify patterns and emerging problems. There have also been some attempts to use transfer functions and cross-spectrum analyses to relate multiple time series.

Recognizing the inadequacy of dynamic models with such incomplete data, attention inevitably turned to the use of equilibrium methods. The commercial fishery had provided the platform for obtaining samples for use in age, growth, and reproductive studies, and for obtaining data on the age or length composition of the catches. These data were used to derive yield per recruit and spawning biomass per recruit curves, while estimates of total mortality were obtained from analysis of age and length-based catch curves. The primary use of these data was to set appropriate minimum legal lengths, at or above the length at maturity, in anticipation that such controls might ensure that each female fish has the potential to spawn prior to capture.

The introduction of recreational fishing licences for some species, such as the western rock lobster (Panulirus cygnus) and abalone (Haliotis spp.), provided both a measure of participation by recreational fishers and a target population for statistical surveys. Regular mail surveys were conducted of some licensed recreational fishers, resulting in a time series of estimates of recreational catch that could be used when fitting dynamic models for the fishery. However, the spatial and temporal distribution of recreational catches and recreational fishing effort are not always the same as those of the commercial fishery, particularly in the case of the western rock lobster fishery. While the collection of such recreational data represented a major advance for our stock assessment of some of the estuarine and nearshore fisheries, the resolution and precision of the recreational catch and effort statistics still restrict their use in more detailed analyses of the data.

Our stock assessments have attempted to overcome deficiencies in data quality through the use of strong, but potentially erroneous assumptions. As an example, consider the stock assessments for the barramundi (Lates calcarifer) fisheries in the northwest of Western Australia. For these fisheries, random samples were drawn from statewide estimates of the rates of increase in recreational fishing participation, and were used to generate a time series of random values to represent the relative increase in recreational effort in the fishery through its history and into its future. The index of recreational effort was set to one for the current year. An estimate of the recreational catch was then randomly selected from a uniform distribution representing the range of estimated catches for the current year, which had been subjectively estimated by experienced research scientists. By comparing this estimated recreational catch to the current commercial catch, an estimate of the relative catchability of the recreational sector was obtained. Using this, together with the time series of relative effort levels for the recreational fishery, the time series of recreational catches was calculated. A biomass dynamics model was then fitted to the combination of recreational and commercial catches. The results of the analysis were stored, and the entire process repeated for a large number of trials. The resulting range of estimates was considered to represent the range of alternative scenarios of historical catches and was used to determine the likely outcomes of several different management strategies. Although the resulting assessments are the best currently available for Western Australia’s barramundi fisheries, the results are highly dependent upon the assumptions that were made. The truth is that we have no data on the catch of barramundi that is actually taken by the recreational sector. Assumptions are an inadequate proxy for research data!

The future

While innovative approaches such as that used for the barramundi fishery, or equilibrium approaches such as those used in many per recruit studies, provide useful information for the management of our estuarine and nearshore fisheries, they are unable to identify the precise status of stocks or to allow assessment of the effectiveness of alternative management strategies. Prior to the 1980s, fisheries modelling and stock assessment was focused on major limited entry fisheries and there was little concern for the less heavily exploited finfish stocks of the estuarine and nearshore environments. In those early years, the approaches used to model the finfish stocks were adequate given the levels of exploitation that existed. However, increasing levels of exploitation and growth in participation in recreational fishing are placing greater pressure on fish stocks. This increasing pressure is leading to concern regarding the levels of exploitation for many finfish stocks, and there is an increasing need to ensure that management controls are effective in ensuring the sustainability of those stocks. In Western Australia, there are currently no controls that constrain the growth of the recreational fishing sector, and other controls, such as bag limits, size limits, gear controls and spatial or temporal closures, must be relied upon to constrain exploitation.

Adequate models that allow the assessment of the stocks of Western Australia’s multispecies estuarine and nearshore fisheries are urgently needed. Exploitation is increasing, yet such increase is likely to be unsustainable, resulting in greater demand for appropriate stock assessment advice to balance precaution with optimal resource utilisation. Our models of these fisheries are inadequate and cannot ensure effective management controls. The lack of time series data on total catch and effort has constrained the use of traditional dynamic fisheries models. Furthermore, with an increasing share of the catch in many of the nearshore and estuarine fisheries gradually being re-allocated from the commercial to the recreational sector, the inadequacy of data will increase and commercial fishery statistics will become more imprecise.

It is unlikely that additional resources will become available for more research, yet there is increasing demand from managers for the scientific advice necessary to develop strategies that maintain the fisheries. Clearly, if models are developed using inadequate data, the uncertainty needs to be exposed to ensure that an appropriate level of precaution is considered when management plans are developed. Furthermore, feedback mechanisms should be established to ensure that, should the ecosystem move beyond the limit reference points, exploitation is reduced sufficiently to ensure recovery!

Conclusion

The inadequacy of past data does not preclude the establishment of improved systems of data collection. Systems that ensure continued collection of reliable data that are sufficient to monitor the fishery and provide necessary feedback must be established, particularly in those fisheries in which the recreational sector becomes the sole or major participant. For such systems, consideration may need to be given to the use of research surveys rather than the collection of fishery-dependent but less reliable data. Future stock assessments will need to use the limited data that are available, and to explore the uncertainty associated with those data. Strategies to manage recreational fisheries will need to be developed that will be responsive to the limited feedback (data) that may become available. Closed loop models are one strategy whereby the robustness of management strategies can be tested.

Acknowledgements

Drs Rod Lenanton and Suzy Ayvazian, and Ms Gabrielle Nowara, are thanked for their advice that aided in the development of the barramundi model, and who, together with other scientists and fishery managers at the Department of Fisheries Western Australia, contributed in many discussions to the concepts that are presented above.