Aquaculture/NSW Fisheries

This report was commissioned by NSW Fisheries (in New South Wales, Australia) to provide modelling support for the Estuary General fishery. Analyses based on ages and lengths of fish and catch data (supplied by NSW Fisheries) were done for dusky flathead (Platycephalus fuscus), luderick (Girella tricuspidata), yellowfin bream (Acanthopagrus australis), sand whiting (Sillago ciliata) and sea mullet (Mugil cephalus).

Growth-curves were fitted to data for ages and lengths. These describe the length of a fish in relation to its age. The commonly used von Bertalanffy growth curve was generally quite difficult to fit to the data because there was little information about the length of young fish. Young fish were too small to be commercially harvested and were therefore not well represented in the available data. Errors associated with the fitted growth curves were estimated. Results indicated that the growth models were quite good for all species except for sand whiting. The figure illustrates the data for age-length and fitted growth curves for yellowfin bream (Acanthopagrus australis).

(Click here to see enlarged figure)

Length-at-age data for yellowfin bream (Acanthopagrus australis). Growth curves that relate the length of a fish with age are also drawn. Separate plots are included for both sexes combined and, separately, for males and females. This report was commissioned by NSW Fisheries to provide modelling support for the Estuary General fishery. Analyses based on ages and lengths of fish and catch data (supplied by NSW Fisheries) were done for dusky flathead (Platycephalus fuscus), luderick (Girella tricuspidata), yellowfin bream (Acanthopagrus australis), sand whiting (Sillago ciliata) and sea mullet (Mugil cephalus).

Growth-curves were fitted to data for ages and lengths. These describe the length of a fish in relation to its age. The commonly used von Bertalanffy growth curve was generally quite difficult to fit to the data because there was little information about the length of young fish. Young fish were too small to be commercially harvested and were therefore not well represented in the available data. Errors associated with the fitted growth curves were estimated. Results indicated that the growth models were quite good for all species except for sand whiting. The figure illustrates the data for age-length and fitted growth curves for yellowfin bream (Acanthopagrus australis).

These growth curves were used in yield-per-recruit analyses which are usually used to explore the trade-off between catching many small fish versus catching few larger fish. The recommended safe rate of caching fish (F0.1) was calculated for all species as a function of age-at-first-capture and also as a function of length-at-first-capture. The latter case is much easier to interpret where length-at-age is variable. It relates fishing mortality directly to minimal legal length. For all species, levels of safe mortality increased as length-at-first capture increased. F0.1 analyses were very sensitive to changes in the estimate of natural mortality (M, or the mortality rate in the absence of fishing). Estimates of M were very uncertain, so we recommended that findings be interpreted with care. We did not recommend that our calculated values of F0.1 be used in decision-making, as there are too many limitations with these initial estimates.

Values of F0.1 are not in themselves useful for management unless they can be contrasted with current rates of mortality due to fishing. Fishing mortality can be approximated from estimates of total mortality and natural mortality. Total mortality (Z) was estimated for all species from the catch-at-age distribution, although confidence in these estimates is compromised by a lack of understanding of recruitment of young fish. Variable recruitment will mask changes in total mortality.

No reliable estimates of natural mortality were available. It is unlikely that data to estimate natural mortality will become available in the near future. We therefore recommended that, to manage these stocks, NSW Fisheries continue to explore methods that do not depend on precise estimates of natural or fishing mortality. One strategy will be to set minimal legal lengths so that harvesting has the least practicable impact on the reproductive output of any species.

NSW Fisheries has elected to use commercial catch as a temporary indicator of the state of stocks in the Estuary General fishery. We wanted to test hypotheses about how frequently pre-determined changes in catches were successful in detecting failure of recruitment or large changes in the survival of mature fish in the fishery. It was also of interest how frequently such indicators sent signals that "cried wolf" or indicated that something was wrong with the fishery when nothing unusual had occurred. These analyses required subjective judgements to define the magnitude of change in the fishery that represents a problem and acceptable probabilities of errors from the indicators.

NSW Fisheries should continue the annual collection of data about ages of fish. This will be the most reliable evidence for recruitment failure and, with time, will enable a better understanding of stock-dynamics. Estimates of recreational catches will be required if assessments of biomass of stock are required. The NSW Estuary General finfish fishery is likely to remain "data-deficient" in the foreseeable future. Sustainable management of this fishery will require development of strategies that are able to deal with uncertainty in estimates of biomass, natural mortality, commercial and recreational fishing mortality. We urged NSW Fisheries to identify the immediate threats that face the stocks and the patterns in observable data that would result if such impacts occurred. Appropriate managerial responses to such scenarios must be clearly defined and strategies designed, to manage the fisheries. These responses and strategies should be tested with simulated data to explore their effectiveness and practicality. Length-at-age data for yellowfin bream (Acanthopagrus australis). Growth curves that relate the length of a fish with age are also drawn. Separate plots are included for both sexes combined and, separately, for males and females.

These growth curves were used in yield-per-recruit analyses which are usually used to explore the trade-off between catching many small fish versus catching few larger fish. The recommended safe rate of caching fish (F0.1) was calculated for all species as a function of age-at-first-capture and also as a function of length-at-first-capture. The latter case is much easier to interpret where length-at-age is variable. It relates fishing mortality directly to minimal legal length. For all species, levels of safe mortality increased as length-at-first capture increased. F0.1 analyses were very sensitive to changes in the estimate of natural mortality (M, or the mortality rate in the absence of fishing). Estimates of M were very uncertain, so we recommended that findings be interpreted with care. We did not recommend that our calculated values of F0.1 be used in decision-making, as there are too many limitations with these initial estimates.

Values of F0.1 are not in themselves useful for management unless they can be contrasted with current rates of mortality due to fishing. Fishing mortality can be approximated from estimates of total mortality and natural mortality. Total mortality (Z) was estimated for all species from the catch-at-age distribution, although confidence in these estimates is compromised by a lack of understanding of recruitment of young fish. Variable recruitment will mask changes in total mortality.

No reliable estimates of natural mortality were available. It is unlikely that data to estimate natural mortality will become available in the near future. We therefore recommended that, to manage these stocks, NSW Fisheries continue to explore methods that do not depend on precise estimates of natural or fishing mortality. One strategy will be to set minimal legal lengths so that harvesting has the least practicable impact on the reproductive output of any species.

NSW Fisheries has elected to use commercial catch as a temporary indicator of the state of stocks in the Estuary General fishery. We wanted to test hypotheses about how frequently pre-determined changes in catches were successful in detecting failure of recruitment or large changes in the survival of mature fish in the fishery. It was also of interest how frequently such indicators sent signals that "cried wolf" or indicated that something was wrong with the fishery when nothing unusual had occurred. These analyses required subjective judgements to define the magnitude of change in the fishery that re