Stock Assessment Review Workshop

STOCK ASSESSMENT REVIEW WORKSHOP

1 2 1 1 C.M. Dichmont , M. Haddon , K.M. Yeomans and K.E. Kelly 1Department of Primary Industries SouthernFisheries Centre PO Box 76, Deception Bay, Qld. 4508 Australia

2Department of Fisheries Australian Maritime College P.O. Box 21, Beaconsfield, Tas, 7270

FI SHERIE S RESEARCH & QF:M:A DEVELOPMENT Queensland Fisheries CORPOR ATIO N DEPARTMENT OF �Management Auth.ority PRIMARY INDUSTRIES

FRDC Project No. 98/129 CONTENTS

STOCK ASSESSMENTREVIEW WORKSHOP ...... 1 1. NON-TECHNICALSUMMARY ...... 3 Principalinvestigator: ...... 3 Objectives ...... 3 Non-technical summary...... 3 2. BACKGROUND ...... 5 3. NEED ...... 7 4. METHODS AND RESULTS...... 8 Training ...... 8 Data collection...... 8 The workshop ...... 8 Workshop extension and proceedings ...... 9 5. ACHIEVEMENT OF OBJECTIVES ...... 10 6. BENEFITS ...... 12 7. ACKNOWLEDGEMENTS ...... 13 APPENDIX 1: INTELLECTUAL PROPERTY ...... 14 APPENDIX 2: PROJECT STAFF ...... 14 APPENDIX 3: CHAIRMAN'SCOMMENTS - MALCOLM HADDON...... 15 Summaryand Recommendations ...... 15 The Workshop ...... 16 The Workshop Process...... 16 The Benefits ...... 16 The Disadvantages...... 17 General Statements regarding Stock Assessment at Deception Bay...... 17 APPENDIX 4: WORKSHOP PROCEEDINGS ...... 22

2 1. NON-TECHNICAL SUMMARY

98/129 Stock Assessment Review Workshop

Principal investigator: C.M. Dichmont Address: Department ofPrimary Industries SouthernFisheries Centre PO Box 76 Deception Bay 4508 Queensland Telephone: 07 3817 9560 Fax: 07 3817 9555

Objectives

1. To review and evaluate existing stock assessment and monitoring programs and methods relating to fisheryresources in SE Queensland, with particularreference to easternking prawns, spanner crabs, tailor and saucer scallops. 2. Use existing data on easternking prawns, spannercrabs, tailor and scallops to develop and test alternative assessment techniques. 3. To advise the Queensland Fisheries Management Authority and DPI Fisheries Group ofthe biological status ofthese resources, and make recommendations on futuredirections and/or priorities forresearch and monitoring. 4. To develop appropriate strategies formonitoring and assessment of southern Queensland fisheries resourcesat levels commensurate with the value ofresources and risk of overfishing. 5. To train Queensland based fisheriesscientists in recently developed stock assessment procedures 6. The consultant should report on workshop proceedings as well as limitations or bottlenecks to effective stock assessment in Queensland. The consultant should also offerrecommendations which may alleviate such limitations to stock assessment in Queensland and improve stock assessment procedures in that state.

Non-technical summary

A threeday stock assessment course was presented by Dr Malcolm Haddon ofthe Australian Maritime College and Dr James Scandol ofthe Quantitative Training Unit forFisheries. Techniques such as biomass dynamic and age based modelling were covered.

Thereafter, a Stock Assessment Review Workshop, fundedby the Fisheries Research and Development Corporation (FRDC), Department ofPrimary Industry, Queensland (DPI) and the Queensland Fisheries Management Authority (QFMA), was held in August at the SouthernFisheries Centre, Deception Bay, Queensland. It was convened by stock assessment scientist Ms Cathy Dichmont ofDPI, and facilitated by Dr Malcolm Haddon fromthe Australian Maritime College (AMC). Commercial fishers, recreational fishers,managers and scientists were all represented at the workshop. It should be noted that the workshop only reviewed, analysed and

3 commented on some of the major fisheryresources of southeast Queensland. The resources chosen forreview were determined through a series of Resource Priority Workshops held within DPI. Those workshops concluded that the primary southeast Queensland resources are spanner crabs (Ranina ranina), easternking prawns (Penaeus plebejus), saucer scallops (Amusium japonicum balloti), sea mullet (Mugil cephalops), tailor (Pomatomus saltatrix) and an inshore finfishcomponent of bream (Acanthopagrus spp), whiting (Siliago spp.) (excluding the trawled stout whiting) and dusky flathead(Platycephalus fuscus).

The workshop reviewed and evaluated present stock assessment and monitoring programs in order to investigate and develop alternative assessmenttechniques. A furtherobjective was to make recommendations on the futuredirections and priorities forresearch and monitoring. Also, where possible, the biological status of the resources being investigated was determined.

The resources reviewed were treated on a species by species basis, with each species being allocated a chapter in the Workshop Proceedings. At the end of each chapter, a detailed list of monitoring and research directions and priorities are given. The authors present a major conclusions chapter which summarises some of the main recommendations, but these recommendations should not be interpretedas representing the exclusive list of proposals and time should be taken to read the full list of recommendations at the end of the report on each species.

Currently, thelongest and, in most resources, the only time series available as an index of biomass is catch rate. However, in some species doubts were expressed concerning the validityof using simple catch rate data as an index of stock biomass. This is especially true forspecies such as spanner crab where complex animal behavioural patterns may affectcatch rate and tailor which is highly aggregatory. In cases where theuse of catch rate is more defensible, the series should be standardised to remove factors thataffect catch rate otherthan biomass.

In many of the draftand new Management Plans developed by QFMA, performance indicators and triggerpoints arementioned. In most cases, these indicators remain untested. In the workshop, the spanner crab TACC decision rules were tested using a delay differencemodel and highlighted many aspects of the shortcomings and strengths of the rules. This typeof work should be extended to include other major species e.g. scallops and easternking prawns.

The independent saucer scallops survey conducted in 1997 provided valuable informationto the workshop. Although surveys are expensive, in many cases their usefulnessfar exceeds their cost. It was highly recommended that this particular survey should continue, but also that independent surveys should be extended to other major resources within Queensland.

4 differentlyto scientists within Queensland. Bream and flatheadotoliths can be fairly easily aged, but there was much difficultywith the whiting species.

The resolution of the logbook database was oftena major stumbling block in developing stock assessment models. Much of the data is still entered within the 30- minute square grids, whereas many management rules require at least 6-minute square or direct latitudes and longitudes. Furthermore, in many cases the data can not be resolved to species level. The most chronic of the species investigated was the inshore whiting. Gear type beyond net and trawl etc. is also not recorded even though, forexample, in the net fishery,beach seine gear have extremely different catchability and selectivity functionscompared to tunnel and gill nets.

Modelling techniques used within the workshop were extremely varied, from simple biomass dynamic models, estimation of total mortality from age information, generalised linear models to a full-scalerelative age model using survey, tagging and catch rate data to estimate population size and recruitment indices. The workshop had limited success in estimating finalpopulation parameters to determine the health of the stocks. Mostly this was due to logbook problems (e.g. resolving whiting catches into species), lack of contrast in the data (e.g. eastern king prawns catchrate trends), definingeffort (e.g. the multi-species and multi-endorsed net fishery)or unknown recreational catch (e.g. tailor and the inshore finfishspecies). However, of extreme use were the management questions that were addressed. The benefitsof a winter closure forscallops, the benefitsof opening for a short period the scallop preservation zones to catch senescent animals and testing the spanner crab TACC decision rules are examples. It can be argued, however, that the three greatest benefitsof this workshop were: • the collation of data, which prior to the workshop was either untraceable, on old computers, only on datasheets or in various filesheld by the scientists involved on the project at the time, and • the communication between all interested individuals, be they biologists, stock assessment specialists, fishersetc., concentrating on species of relevance to south-east Queensland in a hands on forum,and • a list of research priorities and monitoring directions forthe future.

2. BACKGROUND Our proposal used the FRDC fundedNT Workshop ("Towards the sustainable use of NT fisheryresources" FRDC 96/158) as a model so as to review the major projects fundedby FRDC and DPI and to develop stock assessment models of the major resources. The workshop also tied in with a QFMA peer driven requirement to closely assess the status of the spanner crab fishery whichis Queensland's only TAC managed fishery,and the tailor fishery,a very important recreational fisherythat is widely believed to be over-exploited.

Queensland's fishingindustry is based on relatively small landings froma wide range of species. Total wharfsidelandings fromthe fisheriesis valued at about $175 m annually. However, as most individual fishstocks contribute only modestly to this figure,fisheries tend to have a multi-species focus. This is reflectedin a complex, multi endorsement licensing system.

5 The responsibility formanaging Queensland's fisherieslies with the Queensland Fisheries Management Authority (QFMA). This organisation is responsible for ensuring the objectives of the Queensland Fisheries Act are met. One such objective relates to the use of fisheriesresources in an ecologically sustainable manner. To achieve ecological sustainable development, the Authority uses a precautionary approach to management. The Authority is required to develop management plans for each fisheryunder its jurisdiction, and does so through Management Advisory Committees (MACs). MACs (and more broadly, the general community) have placed an increased demand forstock assessment and other scientific advice. Some MACs have already established Stock Assessment Groups for specificresources. Present assessment procedures are almost exclusively focussedon fishery-dependent abundance indices derived fromcommercial catch-effortstatistics. Comprehensive daily statistics are available across the entire Queensland commercial fleetvia a compulsory commercial logbook system, which has been in operation forabout 10 years. While this is a relatively short time-series compared to those forsome NorthernHemisphere fisheries, we believe that this should not be a barrier to the progress of serious scientificallybased stock assessment work.

The application of fisherymonitoring and assessment procedures is still fairly new in southern Queensland. Consequently the development of meaningfulassessments for individual species has been limited by the availability of appropriate expertise and has been largely focusedon per-recruit analysis. Given the increasing levels of concern about the status of fishedresources being voiced by MA Cs and individual scientists, there is great demand on those scientists with mathematical and statistical skills. Rational allocation of scarce assessment skills is a matter to be addressed. A process of in-house resource prioritisation was completed by DPI stock assessment biologist Ms C. Dichmont in 1997. Several criteria were used to prioritise the more than 150 resources, including value of catch, size of catch, size of recreational component, level of conflictetc. This process identifiedthe region's top 5 marine resources as being tailor, easternking prawns, spanner crabs, blue swimmer crabs and scallops. As a result, most stock assessment work since this meeting has concentrated on these species. This has alleviated some of the problems of resource allocation, but has exposed stock assessment researchto little peer review in terms of appropriate methodology.

It is important to note that most of the stock assessment and monitoring projects in this region are at or near completion, because they have been reliant on external (mainly FRDC)funding. As a result, furtherfunding in the form of industry-based cost-recovery and further FRDC applications will have to be investigated within the 1998/99 financialyear. These monitoring programs have been fairlybroad-based, both in terms of species covered and techniques used, and will have to become more specificand less resource-intense to achieve industry funding.

The current FRDC fundedstock assessment and monitoring projects which will be completed in 1998/99 in south-east Queensland are a) "Integrated Stock Assessment and Monitoring Program" (FRDC T94/161) which investigates length- and age-based procedures for monitoring the stocks of several important recreational and commercial coastal and estuarine species in southern Queensland.

6 b) "Support of employment of Stock Assessment Biologist" which has assisted DPI in recruiting a qualifiedand experienced stock assessment biologist to the region and c) "Stock Assessment and Management of Spanner Crabs in SE Queensland" (FRDC 95/022) which investigates the biological parameters forstock assessment and provides the platformfor initial stock assessment attempts.

3. NEED

A number of introductory stock assessment and monitoring programs fundedby FRDC will be completed within the next financialyear. Most of this work has been undertaken with little peer review. There is a need for an externaland highly competent stock assessment consultant to review present research and assessment procedures used forfisheries in southern Queensland,and to identifythe direction of future research and monitoring. It is expected that futuremonitoring programs will be resourced through an industry based cost-recovery program. As a result, a narrowing of objectives and species is needed.

There is a need to look closely at the simulation models being developed by scientists formanagement use in south-east Queensland. For example, the State's only output controlled fisheryis subject to a TAC, which is set fromsustainable yield estimates derived frommodels that are rapidly becoming outdated. Several monitoring programs, experiments and simulation models have been attempted with varying success. This work needs to be exposed to peer review as this fisherymoves (probably) to an ITQ management system. These models should be updated and the present status of the resource needs to be examined, particularly in light of the situation in northernNew South Wales as it is a shared resource.

There are a number of data sets on south Queensland fisheriesthat may be extensive enough to warrant furtherin-depth assessment. There is an opportunity forthis work to be completed at the workshop while the expertise is available.

All the MACs are seeking informationon the status of major fishresources for which they have management responsibilities. They are becoming increasingly aware of the need forobjective and scientificallybased assessment work and are slowly becoming aware of the costs and resource requirements associated with assessments. The supply of initial assessment reports fromthe proposed workshop will reinforcethe move towards objective assessment and associated management recommendations.

Fisheries management in Queensland will be based upon Fishery Management Plans. There is a need to incorporateinto these Plans much more robust and well-conceived monitoring and assessment strategies and decision rules than is occurring at present.

Stock assessment skills are scarce within south-east Queensland. An intensive workshop of most major species will include the majority of fisheriesscientists and technicians within the region. This will expose them to stock assessment methods and their data requirements.

7 4. METHODS AND RESULTS

The project contained fourbasic components: 1. Training to expose Queensland scientists to stock assessment techniques, 2. Data collection and collation, 3. The Stock Assessment Review Workshop, and 4. Extension and publication of proceedings.

Training

The training of Queensland based scientists in stock assessment techniques were approached in two ways. A three day stock assessment course was held prior to the workshop. On the first day, Dr James Scandal presented the stock assessment computer based module from the Quantitative Training Unit forFisheries, Sydney. Topics covered were: • Simple population models, • Parameter estimation, • Growth of individuals, • Standardised indices of abundance, • Stratifiedrandom survey designs, • Stock-recruitment relationships and • Biomass dynamic models. On the next couple of days, Dr Malcolm Haddon of the Australian Maritime College lectured on advanced biomass dynamic models and age-based modelling techniques. Scientists fromNorthern and SouthernFisheries Centre were sponsored to attend.

During the workshop, scientists were exposed to experts within the stock assessment field. Several SouthernFisheries Centre scientists had to verbally present their research forpeer review. Each group that was compiled during the workshop always contained a stock assessment expert, a biologist, a person knowledgable with the database and an industry representative. The above formatallowed scientists to expose their newly learnttheoretical knowledge to real lifedata and situations.

Data collection

Emphasis was placed on data collation into a central, easily accessible and well described formthat would be available to anyone in the future. This work had never been dome beforeat this scale. As a result, informationand data was collated from various sources and in some cases was re-entered fromdatasheets. Also, information on an old out of date computer was transferredto more modernPC's. In some cases, data has been lost for various reasons and these situations have also been described. All data resides on a CD within the centre and the data is to be described in metadata formatand placed in the Blue Pages as part of the Coastal Atlas.

The workshop

8 A Stock Assessment Review Workshop was convened by the project Principal Investigator Ms Cathy Dichmont and facilitatedby Dr Malcolm Haddon fromthe Australian Maritime College over a two week period between Sunday 16th August and Thursday 28th August. Commercial fishers, recreational fishers, managers and scientists were all represented at the workshop. The workshop reviewed, analysed and commented on some of the major fisheryresources of southeast Queensland. The resources chosen forreview were determined through a series of Resource Priority Workshops held within the Department of Primary Industries. These are spanner crabs (Ranina ranina), eastern king prawns(Penaeus plebejus), saucer scallops (Amusium japonicum balloti), sea mullet (Mugil cephalus), tailor (Pomatomus saltatrix) and an inshore finfishcomponent of bream (Acanthopagrus spp), whiting (Siliago spp.) ( excluding the trawled stout whiting) and dusky flathead (Platycephalus fuscus).

The workshop sessions for each of these species and species groups had the same formatand each session lasted between 1.5 and 2 days. The focusof the workshop was upon the stock assessment of each species and how this could be used to improve on management advice.

The first step was always an introduction to the biology of the species, its current management arrangements, the data available and the research that had already been conducted. From this exposition Dr Haddon, with the groups agreement and prompting, would propose four or fiveactivities which could likely be profitably pursued by smaller sub-groups. The members of the workshop would then be sub divided among these activities and each such sub-group would work on their particular problem forbetween four and six hours. Each sub-group would then report back to the complete workshop andfrom that and the ensuing discussion the final recommendations and suggestions of the workshop would be devised and agreed upon. This formatwas surprisinglysuccessful and it is the opinion of the authors that the suggestions produced were a consensus of opinion.

Workshop extension and proceedings

The day afterthe workshop, an open seminar series was held within the centre. More than 50 people attended, mostly fromthe trawl and finfishcommercial sectors, QCFO and Sunfish. The progress and results of each species group was presented in a half hour seminar of which, 5 minutes of questions were allowed. Thereafter, a freeBBQ lunch was provided to furtherdiscussions and interaction.

Afterthe workshop, the group's models were in various stages of completion. Any incomplete work was finalised by Ms Dichmont and Dr Haddon. Furthermore, all results were checked. A Stock Assessment Review Workshop Proceedings was written in which all the work undertaken in the workshop was described (Appendix 4). Final recommendations were also collated therein. Transcripts of the discussions during the talks, the group reports and futureresearch direction sessions were made. These are included as a disk in the Workshop Proceedings.

Results and recommendations of the workshop were reported to the relevant Management Advisory Councils. An article was published within DPI Prime News and Queensland Fisherman. Recommendations of researchdirections were made by

9 Dr Malcom Haddon to QFIRACfor the next round ofFRDC project proposals. Ofthe four projects supported,two were subsequently funded.

5. ACHIEVEMENT OF OBJECTIVES

Objective 1. To review and evaluate existing stock assessment and monitoring programmes and methods relating to fisheryresources in SE Queensland, with particular reference to eastern king prawns, spanner crabs, tailor and saucer scallops.

All available data collection and stock assessment research was reviewed within the workshop. Species coverage was increased to include spanner crabs, saucer scallops, easternking prawns,sea mullet,tailor and an inshore finfishcomponent offlathead, bream and whiting. The main method ofreview was through presentations following the general formatof: • Biology and world-wide distribution (presented by the relevantbiologist) • Description ofthe fishery(presented by a fisher), • Present andfuture management (presented by a QFMA manager), • Available data (presented by a scientists most familiarwith the data), • Present analysis and modelling (presented by a stock assessment modeller that has worked on the species), and • Present monitoring program and its cost.

Recommendations were covered in sessions facilitatedby Dr Haddon entitled: • Discussions ofpossible models that could be developed, • Explanation ofmodels used/suggested in the workshop,their assumption and data needs,and • Discussion offuture monitoring program and possible stock assessment models.

Objective 2. Use existing data on eastern king prawns, spanner crabs, tailor and scallops to develop and test alternative assessment techniques.

All available data was collated for the workshop, which included spanner crabs, saucer scallops, easternking prawns,sea mullet,tailor and an inshore finfish component offlathead, bream and whiting. Available data was presented and summarised for the participants so that fulluse ofthe data could be made. In the case of easternking prawns,monthly catch rate data fromNew South Wales and in spanner crabs, taggingdata fromNew South Wales's studies,were obtained. For the recreational species discussed,data from QFMA'srecreational survey was presented and analysed. The Australian National SportfishingAssociation tagging data and recreational club catch rate was also included.

Stock assessment models were successfully developed foreastern king prawns and saucer scallops. A monthly biomass dynamic model for easternking prawn estimated recruitment levels forthe New South Wales and Queensland resource. Unfortunately, the model could not estimate present biomass levels relative to virgin levels due to the lack ofcontrast in the data. A relative age model that incorporates catch rate,survey and tagging data was developed for scallops. A recruitment index was estimated in

10 which the relatively low level of recruitment over the last fewyears can be observed. Modelling in the spanner crab section concentrated on the application and standardisation of catch rate and on developing a single growth curve incorporating all available dredge, tagging and size information. Sea mullet models concentrated on estimating total mortality and growth. Although estimates are provided, uncertainty in the estimates is high due to the short data series. The most important issue addressed by the tailor groups was directly or indirectly linked to the extremely high estimates of natural mortality obtained fromcatch-at-age curves. Several arguments against the conclusion that the resource is overexploited can be given. Most commonly voiced are that the ageing is incorrect and that large animals move offshoreand thereby become unavailable to the fisheryand bias the sample. In combination with the WesternAustralian scientists invited, several catch curves were produced. These investigated thebias and variance between readers fromQueensland and WesternAustralia, between readers in Queensland and between multiple readings from1a single scientist. Total mortality values could thereforerange from0.8 to 2 year· depending on interpretation of otolith rings. The inshore finfishsessions were dominated by the lack in resolution of the data in terms of species, method of capture and location. In some cases, algorithms were developed to separate the catch by method or species, but in the case of whiting only rules of thumb could be developed.

Objective 3. To advise the Queensland Fisheries Management Authorityand DPI Fisheries Group of the biological status of these resources, and make recommendations on future directions and/orpriorities for research and monitoring.

The results of the proceedings have been presented to the relevant MACs within QFMA, being Crab, Trawl and Subtropical FinfishMAC. Furthermore, a report back seminar in which each section was given a 30-minute presentation was completed in the Centre, in which fishers,managers and executives fromQFMA and DPI attended. The results of the workshop were also published in DPI Prime News and The Queensland Fisherman. Dr Malcolm Haddon wrote a recommendation forproject research priorities to QFIRAC in which two of the fourprojects supported have subsequently already been funded.In the proceedings are clear sections recommending futureresearch and monitoring directions and priorities.

Objective 4. To develop appropriate strategies for monitoring and assessment of southern Queensland fisheriesresources at levels commensurate with the value of resources and risk of overfishing.

In each species section, a half-day session was dedicated to the discussion of monitoring and assessment priorities and research needs within southeast Queensland. As a result, in the proceedings, a heading in each resources section is dedicated to the appropriate strategies that should be taken both in the long and short term. Dr Malcolm Haddon wrote a recommendation for project research priorities to QFIRAC in which two of the fourprojects supported have already been funded.

11 Objective 5. To train Queensland based fisheries scientists in recently developed stock assessment procedures

As stated in the section of training within Methods, the training of Queensland based scientists in stock assessment techniques were approached in two ways. A three day stock assessment course was held prior to the workshop. On the firstday, Dr James Scandol presented the stock assessment computer based module fromthe Quantitative Training Unit forFisheries, Sydney. On the next couple of days, Dr Malcolm Haddon of the Australian Maritime College lectured on advanced biomass dynamicmodels and age-based modelling techniques. Scientists fromNorthern and SouthernFisheries Centre were sponsored to attend.

During the workshop, scientists were exposed to experts within the stock assessment field. Several SouthernFisheries Centre scientists had to verbally present their research forpeer review. Each group that was compiled during the workshop always contained a stock assessment expert, a biologist, a person knowledgable with the database and an industry representative. The above formatallowed scientists to expose their newly learnt theoretical knowledge to·real lifedata and situations.

Objective 6. The consultant should report on workshop proceedings as well as limitations or bottlenecksto effective stock assessment in Queensland The consultant should also offerrecommendations which may alleviate such limitations to stock assessment in Queensland and improve stock assessment procedures in that state.

Appendix 3 is a document by Dr Haddon, which has also been given to QFMA and QDPI in which any bottlenecks and limitations to effectivestock assessment in Queensland are discussed. Furthermore, the consultant has reported on the workshop and proceedings. It should be stated that the consultant, Dr Haddon, contributed enormously to the success of the workshop.

6. BENEFITS

This project consolidated the direction of future stockassessment and monitoring programs and is of high regional significance.

The project has in part or fullproduced the followingbenefits: • A review of present stock assessment and monitoring research, • A strategic plan forstock assessment and monitoring within the SE region, • The status of major fisheryresources or proposed research direction to achieve fullassessment of the resource, • Exposure of stock assessment techniques and their data needs, • Stock assessment input into Fishery Management Plans, • Peer review of mainly FRDC-based projects, and • More robust advice to QFMA.

12 The major beneficiaries are the QFMA MACs who are responsible forthe development of Management Plans. In the long-term, Management Plans based on objective assessment will be of major benefitto the commercial and recreational fishingsectors. Participating scientists and QFMA fisheriesmanagers will benefits fromexposure to stock assessment techniques and both QFMA and QDPI will benefit through the development of strategic research priorities.

7. ACKNOWLEDGEMENTS

The Fisheries Research andDevelopment Corporation (FRDC), Department of Primary Industries (DPI) and Queensland Fisheries Management Authority (QFMA) fundedthis workshop. It was a major undertaking forseveral people, however, the largest contributions came fromDr Malcolm Haddon, workshop facilitator, Proceedings author and editor; Ms Kath Kelly, workshop coordinator and transcriptor of the audio tapes; Ms Kate Yeomans, data collector, collator and Proceedings editor. Your dedication is much appreciated. I would like to thank Dr Ian Brown and Mr Bob Pearson for convincing the various parties and interest groups that this workshop was essential and forsecuring financingfrom QFMA and DPI. Each session had an allocated coordinator, which really reduced my workload. They were; Dr Ian Brown forspanner crabs, Dr Tony Courtney for easternking prawns, Mr Mike Dredge for scallops, Mr Ian Halliday formullet and Mr Simon Hoyle for tailorand the bream/whiting/flatheadsessions. Several scientists have reviewed the manuscript, but I would like to make a special mention of Mr Mike Dredge, Dr Ian Brown and Dr Daryl McPhee. I would also like to thank the laboratory forallowing me to use the tearoom and other facilitieswithout complaint. Most of all, I really appreciate the help fromthe administrative stafffor what was a difficult and diverse project.

13 APPENDIX 1:INTELLECTUAL PROPERTY

Based on the original agreement, FRDC' s proportion of ownership of intellectual property, decided on financialcontribution, is 45.1%. Proceedings of the workshop have been included in this project but will also be published independently as a document on its own. Results have been published in an industry journal. Some work has already been taken further, mostly by other FRDC projects. These are the effortstandardisation components described in FRDC T94/161 and FRDC 95/022. It is the intention that other components may be written up in peer-reviewed journals. No commercial patents are expected fromthis project.

APPENDIX2:PROJECTSTAFF

Project staff Name Organisation Principal Investigator Ms C.M. Dichmont DPI Queensland Workshop facilitatorand DrM. Haddon Australian Maritime College consultant Workshop coordinator MsK.E. Kelly DPI Queensland Co-investigators Dr I.W. Brown DPI Queensland (alphabetically) Dr A.J. Courtney MrM. C.L. Dredge Mr I. Halliday Mr S. Hoyle MsK.M. Yeomans

Apart fromthe above scientists that contributed to all facetsof the project, more than 45 people participated in the workshop itself. DrsJames Scandol andMalcolm Haddon lectured during the three-day stock assessment course.

14 APPENDIX 3: CHAIRMAN'S COMMENTS - MALCOLM HADDON

Stock Assessment Review, Southern Fisheries Centre, Deception Bay. 1 ih and 28th August 1998

Summary and Recommendations

• The workshop, with its particular format, worked exceedingly well as a source of recommendations foridentifying issues or importance and futureresearch needs. It also acted as a detailed formof extension of ideas about stock assessment and permitted detailed discussion on all sides of the issues raised.

• The two weeks of workshops were very intensive and if futureworkshops are organised they should be limited to a maximum of a single week. The amount of work involved in preparing forthe two weeks was onerous forthe already busy staffof the SouthernFisheries Centre.

• The present commercial fisheriesdatabase is inadequate in manyways for the production of reliable stock assessments. The database itself has many errorsand problems in relation to key stocks. With some urgency, the QFMA need to address the problems with the database. The data already available needs cross checking, the quality control on data entry needs to be improved, and the fisheries scientists who are expected to use the data to provide management advice should be consulted about the structural changes needed to the database and the reporting forms.

• The feasibilityof conducting fishery independent surveys on suitable species, such as saucer scallops, spanner crabs, and Tailor, should be investigated to discover workable ways of funding such enterprises.

• An obvious threat to sustainability for some species, such as easternking prawn and sea mullet, is the factthat these stocks extend and migrate across State boundaries. Until all States share a common set of management objectives for such species they will be vulnerable. Strategies need to be developed for improving and formalisinginter-State communication with respect to shared or migratory species.

• The performance indicators being proposed for use in many of the Queensland fisheriesconsidered were of varied nature but there were many which used a catch rate or assumed associated biomass relative to some earlier biomass level. Such performance indicatorshave their problems and need furtherinvestigation and development. In the workshop a modelling procedure was developed to test the efficiency of a set of such indicators. This approach should also be pursued further as being of general use across Australia. The relationship between fishery management objectives and their associated performanceindicators needs to be tighter wherever possible.

15 The Workshop

The Executive Summary provides a description of the origin and development of the Stock Assessment workshop run at Deception Bay over the two week period Monday 1h th 1 ? to Friday 28 August 1998.

The species and species groups considered by the workshop were Spanner Crabs, EasternKing Prawn, Saucer Scallops, Sea Mullet, Tailor, and Bream/Whiting/Flathead. Particular detailsof the recommendations foreach are included in the body of the text of the main report and the executive summaryand will only be revisited here to illustrate particular points.

The Workshop Process

The workshop sessions foreach of the species and species groups all had the same formatand each session lasted between 1.5 and 2 days. The focusof the workshop was upon the stock assessment of each species and how this could be improved and, in turn, used to improve on management advice.

The firststep of the process for each species or species-group was always an introduction to its biology, the operational practices of its fishery, the data available, the research that had already been conducted, and the current management arrangementsfor the fishery. From this exposition, input fromworkshop participants, and more detailed reading, I would formulate, with the groups agreement and prompting, four or five areas of analysis which were thought of as good candidates for providing an indication of future directions or even an answer to a particular research problem. The members of the workshop would then be organised into sub-groups and each would work on one of the particular problems for between four and six hours.

Each sub-group had members of all sectors represented at the workshop. They would firstgather the data or informationthey required and then collect in a separate room or part of the main meeting room and push the work forwardin a collaborative manner. Aftertaking their particular problem as far as they could in the time available, each sub-group would report back to the complete workshop fordiscussion and commentary.

After all sub-groups had reported back to the workshop a discussion always followed during which the finalrecommendations and suggestions about issues of concern and futureresearch priorities would be devised and agreed upon forthe species or species group concerned. An objective of these discussions was to obtain a consensus but where that was not possible the differentpoints of view were described.

The Benefits

1. The full-workshop/mini-workshopapproaches, and the particular format used, were surprisingly successfulat generating results of either immediate value or indicative value forfuture stock assessments. The discussions followingthe mini workshops were oftenuseful in clarifyingissues considered important by the fishingindustry representatives, clarifyingmanagement objectives, and

16 identifyingactions which may improve the assessment and management of the different fisheries.

2. The workshop automatically acted as an extension process that described and explained many stock assessment methods to Industry and Management representatives. There was an exchange of ideas among all members of the workshop.

3. The open discussion and the opportunities given to everyone who wanted to contribute meant that the process was sometimes slower than was always strictly necessary. However, I believe everyone would agree that the suggestions and recommendations produced were a consensus of opinion, as faras that was possible.

4. The progress of the analyses in the mini-workshops also appeared slower than they could have been but by having the sub-groups made up of both scientists and fishers both sides oftenbenefited from trying to explain to the other their perceptions of each fishery.

5. Preparing forthe stock assessment workshop forcedstaff at the SouthernFisheries Centre, Deception Bay, to collate all available data and other informationand raised their awareness of what historical informationwas available.

The Disadvantages

Preparing forthe two weeks of stock assessment workshops forcedstaff at the SouthernFisheries Centre, Deception Bay, to spend much time in gathering and organising informationinto easily useable computer files. The step of organising the available informationis undoubtedly a good thing but it exposed already busy people to furtherstress and aggravation. Undertakinga workshop with such an ambitious agenda is something that should not be undertaken lightly.

General Statements regarding Stock Assessment at Deception Bay.

Past Data Collection, its Analysis and Interpretation

Many Australian fisherieshave a relatively low landed value and yet the size of the country and its coastline means that performingany usefulassessment on the status of a fish stock can prove to be both difficult and expensive. The relatively low value means that the number of fisheriesscientists also tends to be relatively low. It would be fairto say that at all fisheriesresearch centers in Australia, and the Southern Fisheries Centre at Deception Bay is no exception, the scientificstaff tend to be stretched to their limit in terms of workload.

The fisheries research carried out at the SouthernFisheries Centre invariably appears to have been executed as well as available equipment and fundingwould permit. There is a history of staffat the SouthernFisheries Centre formally publishingat least some of their studies of the biology of commercial species. Longer term stock assessment and monitoring has not fared so well although occasional, localised, population surveys of species such as saucer scallops were conducted. Voluntary

17 logbook schemes were devised forselected fishersfor a number of species and this was the start of attempts at stock monitoring. Such data proved very usefulduring the workshop process when discussing fisheryperformance now relative to earlier dates.

The Commercial Catch and EffortDatabase in Queensland

Every State in Australia appears to have its own set of fisheriesassessment and management problems. To a large extent, Queensland's problems, in termsof its fisheries,derive fromearlier attempts to encourage the development of the Australian fishingfleet. By the time restrictions on new fishing licenseswere instigated in Queensland there were over 800 trawl licenses and hundreds of other licenses in the various fisheries.The logistic problems involved in collecting meaningful fisheries data fromsuch a large set of fishersare enormous. This is reflectedin the factthat except fora very fewfisheries thereare fewuseful fisheries statistics prior to 1990 or even later.

The current logbook scheme, while a great improvement over earlier schemes (especially when no data other than landings were collected), is still inadequate for many stock assessment purposes. Many of the fisheryperformance indicators suggested by the draftfisheries management plan, and forthe individual fisheries, use catch-effortdata in a comparison with earlier catch rates. Without proper error checking when the data are input, and without various design flawsbeing rectified, the proper functioningof the performance indicatorswould be compromised. One big problem is the poor resolution of species identification. For example, the use of "Bay Prawns" indicates a mix of prawn species by size rather than by species. A further example of a major problem is the mixing of ways of reporting catch in the scallop fishery.Catch is reported as both baskets and kgs weight, but sometimes only one is provided and then it can fallinto the wrong database column (e.g. a 1,000 kg catch can be mis-labeled a 1,000 basket catch). Such problems make using the current database difficult.Potentially the errors could be very large making any stock assessments very confused.

The current commercial fisheriesdatabase of catch and effortinformation has some general uses. However, the problems which exist in relation to data quality, missing information,and confuseddata (species lumped together, search times not recorded, failureto separate efforttypes, mixing of catch reporting, etc, - see main report) mean that stock assessments based on this informationcould be flawedfundamentally. Under these circumstances the options are forscientists to introduce extra, voluntary log-book schemes or, formore detailed work, to conduct fisheryindependent surveys.

In Queensland, fishery independent research and monitoring has been patchy at best so there has been no consistent long term stock monitoring. However, there is a history of relatively continuous work on scallops, easternking prawn, and spanner crabs. These works, however, are a reflection of the effortsof individual scientists or groups of scientists, not necessarily as a result of a conscious policy except an implicit one of investigating important fisheries.The intermittent nature of this work relates to a number of factors.First, the patchy history of the collection of centralised fisheries statistics, second, to the intermittent nature of fundingfor fisheries research, especially by way of setting in place routine stock monitoring, and third, the

18 opportunities that arise for particular fisheriesscientist to carryout assessment work on particular species.

With the relatively recent explicit adoption of sustainability as a key management objective there has been an increased emphasis towards research into long term monitoring of stock status. To expect there to have been more by way of stock monitoring in the past would be unjust.

Current Fishery Performance Measures and Reality

The stock monitoring or performance indicators, or both, forspanner crabs, eastern king prawns, and saucer scallops, all rely heavily on commercial catch effortdata. Apart fromthe many problems which exist with this database, many people at the workshop expressed doubts about the validity of using such informationto access stock performance;it was thought that simple catch effortdata was not a good indicator of relative stock size. Thus, it is possible to use the current performance indicators but whether they reliably reflectnature is questionable.

The alternativesuggested was, naturally, fisheryindependent surveys. The best examples of the value of such surveys were provided by the stratifiedrandom surveys forsaucer scallops. During the workshop, the availability of the informationgained fromthe first surveywas used fora number of purposesand led to suggestions for adding up to $3 million of value to the fisherythrough the use of seasonal and spatial closures. To adequately manage a resource such as saucer scallops, which can have highly variable recruitment, such surveys are possibly the only viable option that could lead to optimal sustainable management.

Fishery Independent Surveys

The obvious disadvantage of such fisheryindependent surveys is that they are expensive to run. Such surveys have been suggested forsaucer scallops, spanner crabs, and Tailor. The design of such surveys would be no problem with the present staffat the Southern Fisheries Centre.For example, the previous scallop surveys were nicely designed to cover the ground and provide the needed informationin a valid way.

The fundingof such surveys would be the biggest problem. Many options were discussed in the workshop ranging froma marine recreational fishinglicense, a furtherIndustry levy, a system of charter for catch or effortrights, and combinations of such approaches. Which approach would work best in Queensland will probably vary by fishery.

The workshop demonstrated that fora fewfisheries, such as Tailor, the recreational catch was higher than the commercial catch. Recreational fishers appeared not to be over troubled by the idea of a marine recreational license as long as it was guaranteed that any funds raisedwould be used forrecreational fisheriesresearch and compliance.

Whichever method for fundingthe fisheryindependent surveys is selected, the importance of such surveys, properly executed, can be tremendous. However, not all

19 species are suited to such surveys. Just as with commercial catch-effortdata, care must be taken to restrict such analyses to those species (such as scallops, spanner crabs, and Tailor) which might provide results ofsufficient statistical precision to be usefulas a basis formanagement advice.

Other General Problems

The EasternKing Prawnand Sea Mullet fisheriesprovide examples of where single stocks straddle the waters ofmore than a single State. As elsewhere, a disagreement between States, and sometimes the Commonwealth, over the management of fisheries resources sometimes acts against successfulcollaboration on stock assessment. The Deception Bay workshop was, fora short time, under threat of scientists fromNSW not being allowed to attend. Fortunately, in Australia the community of fisheries scientists is relatively small and there is a history of a good level of collaborative work between people in different States.

The argument forthe separation ofthe management from the assessment offisheries is supported by such difficulties and will not be rehearsed here. Things would become easier forstock assessment, and hence sustainability, when all States have consistent management forcross-border species. EasternKing Prawn and Sea Mullet aretwo species which could be vulnerable to assessment in one State not accounting forall that happens to the stock in other States.

Performance Indicators

Performanceindicators (PI) being suggested for the management of various fisheries in Queensland are oftenbased upon catch-effortdata. These types ofPI are dangerous if, in fact, there is no relationship between the stock status and catch-rates.

There are a number ofquestions about performance indicatorsthat should be addressed and these include such basic ones as whether performanceindicators are an efficientway of managing a fishery. Certainly, when catch rates are used as limit measures of performanceit must be recognised that we are only dealing with estimates ofrelative catch rate. We can thus never be 100 % certain that a trigger point has been reached. Any performanceindicator which utilises a single number as a threshold would sufferfrom the same criticism.

Some PI measures are better than others but currently there are no agreed upon criteria fordistinguishing good frombad. In the workshop a modelling procedure was used to test the efficiency of a set ofproposed Pls which acted upon threshold catch rate levels being reached. This modelling test procedure worked well and should be pursued furtheras a means of comparing differentsets of performanceindicators.

It is ofconcern that Pls that use either catch rates or their assumed stock biomass levels relative to some selected referenceyear have an arbitrary nature. Such indicators or thresholds are difficultto relate to any specificmanagement objective. Although the link can be made to sustainability ifthere is some notion ofthe biomass required to ensure reliable levels of recruitment (not known for any Australian fishery

20 except empirically through comparison with seemingly equivalent fisheries elsewhere).

In the development of quantitative performance indicators (catch rates, economic performance, geographical extent of the fishery,etc) it is necessary that single year effectsare not over-emphasised. There is much remaining to be done in the design and use of performanceindicators when managing a fishery.In the workshop there was active discussion of performanceindicators and their strengths and weaknesses. There is also active discussion continuing in a number of the stock assessment groups operating around the country, especially in the South East Fishery groups. Given Australia's and especially Queensland's extensive use of such management devices this discussion and development should be encouraged and monitored.

21 APPENDIX 4: WORKSHOP PROCEEDINGS

Attached are the Workshop Proceedings. Copies have been professionallyprinted and areavailable as a separate document (ISSN Number 0728-067X). A printed copy will be sent to each participant.

Southern Fisheries Centre Deception Bay, Queensland 16 - 28 August 1998

Proceedings of the South-East Queensland Stock Assesstnent Review Workshop

Editors C.M. DICHMONT, M. HADDON, K. YEOMANS ANDK. KELLY

FISHERIE S RESEARCH & QFM:A DEVELOPMENT Queensl.:lnd Fis)'lcrie:;. CORPORATIO N fHPARTMENT OF Manage:rnent Authority "'-,.,.�·r"'"·"'-..,,_ ,,•.i"''·"' •w"""'""'·½ ..,, ½ v. PRIMARY Conferenceand Workshop Series QC99003 ISSN 0728-667X Agdex 486

Information contained in this publicationis provided as general advice only. For application to specific circumstances, professional advice should be sought.

The Department of Primary Industries, Queensland has taken all reasonable steps to ensure that the information contained in this publication is accurate at the time of production. Readers should ensure that they makeappropriate inquiries to determine whether new information is available on the particular subject matter.

Copyright protects this publication. Except for purposes permitted by the Copyright Act, reproduction by whatever means is prohibited without prior written permission of the Department of Primary Industries, Queensland. Inquiries should be addressed to:

Manager, DPI Publications Department of Primary Industries GPO Box46 Brisbane Qld 4001 Contents

Executive summary

1. INTRODUCTION 1 1.1. Workshop Introduction 3 1.2. Authorship Note 4 1.1.1. Chapter 2. Spanner Crab 4 1.1.2. Chapter 3. Eastern King Prawns 5 1.1.3. Chapter 4. Saucer Scallop 5 1.1.4. Chapter 5. Sea Mullet 5 1.1.5. Chapter 6. Tailor 5 1.1.6. Chapter 7. Bream, Whiting, Flathead 5 1.1. 7. Chapter 9. Available Data 5

2. SPANNER CRABS 7 2.1. Participants 8 2.2. Introduction 9 2.3. Spanner Crab Workshop Tasks 11 2.3.1. Effortstandardisation 11 2.3.2. Develop a yield-per-recruit model to investigate evidence of growth oveifishing 12 2.3.3. Investigation of the application of a production/yield model 14 2.3.4. Develop a simulation model to test the robustness of a set of Decision Rules 20 2.3.5. Examine all existing growth and mortalitydata 25 2.4. Conclusions 26 2.4.1. Group reports 26 2.4.2. Overall progress to date 27 2.4.3. Monitoring, Research Direction and Priorities 28 2.5. References 29

3. EASTERN KING PRAWNS 31 3.1. Participants 32 3.2. Introduction 33 3.2.1 Biology 33 3.2.2 Description of Fishery 36 3.3. EasternKing Prawn Task Groups 39 3.3.1 Characterisationof Catch Rates 39 3.3.2 Distribution of effort 41 3.3.3 Evidence or arguments for recruitment and/or growth oveifishing 43 3.3.4 Establishment of a recruitment index through simulation models 45 3.3.5 Location and size distribution 41 3.4. Conclusion 50 3.3.6 Group reports 50 3.3.7 Progress to date 51 3.3.8 Monitoring, Research Direction and Priorities 52 3.5. References 54 4. SAUCER SCALLOPS 57 4.1. Participants 58 4.2. Introduction 59 4.3. Saucer Scallops Workshop Group Tasks: 60 4.3.1. Value and cost of Preservation 'Zones 60 4.3.2. Value of a potential winter/spring closure 62 4.3.3. Effortstandardisation 64 4.3.4. Stock assessment 66 4.4. Conclusions 72 4.4.1. Group reports 72 4.4.2. Progress to date 73 4.4.3. Monitoring, Research direction and Priorities 74 4.5. References 75

S. SEA MULLET 77 5.1. Participants 78 5.2. Introduction: Sea Mullet in Queensland and New South Wales 79 5.2.1. FisheryStatistics 79 5.2.2. Value of the Fishery 80 5.2.3. Components of the Queensland Fishery 80 5.2.4. Ageing 81 5.2.5. Size and age structure 81 5.2.6. Reproduction 84 5.2.7. Movement 85 5.2.8. Assessment of the legal size 85 5.3. Mullet Task Groups 86 5.3.1. Discussion of Effortin Queensland 86 5.3.2. Examination of catch records 87 5.3.3. Jnvestigation of growth and ageing information 89 5.4. Conclusions 93 5.4.1. Group reports 93 5.4.2. Progress to date 94 5.4.3. Monitoring, Research direction and Priorities 94 5.5. References 96

6. TAILOR 97 6.1. Participants 98 6.2. Introduction 99 6.2.1. Growth 99 6.2.2. Mortality 100 6.2.3. Fishery 102 6.2.4. Yield-per-recruit 105 6.3. Tailor Task Groups 106 6.3. l. Size structure and distribution 106 6.3.2. Estimating relative commercial and recreational harvest 108 6.3.3. Analysis of recreational and commercial catch rates 108 6.3.4. lnvestigation of ageing and growth estimates 111 6.4. Conclusions 113 6.4.1. Group reports 113 6.4.2. Progress to date 114 6.4.3. Monitoring, Research direction and Priorities 114 6.5. References 116 7. BREAM, WIDTING AND FLATHEAD 119 7 .1. Participants 120 7 .2. Introduction 121 7.2.1. Description of the fishery 121 7 .2.2. Development of the fishery 121 7.2.3. Management 125 7 .2.4. Bream 126 7 .2.5. Whiting 130 7 .2.6. Flathead 134 7.3. Bream, Whiting and FlatheadTask Groups 136 7.3.1. Investigation of catch rates from CFISH 137 7.3.2. Estimate relative commercial and recreational catch 138 7.3.3. Discuss and analyse the usefulness of the recreational tagging database 142 7 .3.4. Re-examine growth rate estimates 143 7.4. Conclusions 147 7.4.1. Group reports 147 7.4.2. Progress to date 148 7.4.3. Monitoring, Research direction and Priorities 149 7 .5. References 151

8. SUMMARY OF MAJOR RESEARCH PRIORITIES 153 8.1. Spanner Crab Monitoring, Research Direction and Priorities 155 8.2. EasternKing Prawn Monitoring, Research Direction and Priorities 155 8.3. Saucer Scallop Monitoring, ResearchDirection and Priorities 156 8.4. Sea Mullet Monitoring, ResearchDirection and Priorities 156 8.5. Tailor Monitoring, Research Direction and Priorities 157 8.6. Bream, Whiting and Flathead Monitoring, Research Direction and Priorities 157

9. AVAILABLE DATA 159

10. ACKNOWLEDGEMENTS 169 11. LIST OF PARTICIPANTS 171

12. COMPUTERSOFTWARE UTILISED 175

13. ABBREVIATIONS 177

14. TRANSCRIPTS 179 Executive Summary

A Stock Assessment Review Workshop, funded by the Fisheries Research and Development Corporation (FRDC), Department of Primary Industry, Queensland (DPI) andthe Queensland Fisheries Management Authority (QFMA), was held in August at the Southern Fisheries Centre, Deception Bay, Queensland. It was convened by stock assessment scientist Ms Cathy Dichmont of DPI, and facilitated by Dr Malcolm Haddon from the Australian Maritime College (AMC). Commercial fishers, recreationalfishers, managers and scientists were allrepresented at the workshop. It should be noted that the workshop only reviewed, analysed and commented on some of the major fisheryresources of southeast Queensland. The resources chosen for review were determined through a series of Resource Priority Workshops held within DPI. Those workshops concluded that the primary southeast Queensland resources are spanner crabs (Ranina ranina), eastern king prawns (Penaeus plebejus), saucer scallops (Amusium japonicum balloti), sea mullet (Mugil cephalops), tailor (Pomatomus saltatrix) and an inshore finfishcomponent of bream (Acanthopagrus spp), whiting (Siliago spp.) (excluding the trawled stout whiting) and dusky flathead (Platycephalusfuscus).

The workshop reviewed and evaluated present stock assessment and monitoring programs in order to investigate and develop alternativeassessment techniques. A further objective was to makerecommendations on the future directions and priorities forresearch and monitoring. Also, where possible, the biological status of the resources being investigatedwas determined.

The resources reviewed were treated on a species by species basis, with each species being allocated a chapter in theWorkshop Proceedings. At the end of each chapter, a detailed list of monitoring and research directions and priorities are given. The authorspresent a major conclusions chapter which summarises some of the main recommendations, but these recommendations should not be interpreted as representing the exclusive list of proposals and time should be taken to read the full list of recommendations at the end of the report on each species.

Currently,the longest and,in most resources, the only time seriesavailable as anindex of biomass is catch rate. However, in some species doubts were expressed concerning the validity of using simple catch rate data as an index of stock biomass. This is especially true for species such as spannercrab where complex animal behavioural patterns may affectcatch rate and tailorwhich is highly aggregatory. In caseswhere the use of catch rate is more defensible,the series should be standardisedto remove factors that affectcatch rate other thanbiomass.

In manyof the draftand new Management Plans developed by QFMA, petforrnance indicators and trigger points are mentioned. Inmost cases, these indicators remain untested. In the workshop, the spanner crab T ACC decision rules were tested using a delay differencemodel and highlighted many aspects of the shortcomings and strengths of the rules. This type of work should be extended to include other major species e.g. scallops and easternking prawns.

i The independent saucer scallops survey conducted in 1997 provided valuable information to the workshop. Although surveys are expensive, in many cases their usefulnessfar exceeds their cost. It was highly recommended that this particular survey should continue, but also that independent surveys should be extended to other major resources within Queensland.

Age data from otoliths were of varyingvalue depending on the species concerned. For sea mullet, even though the data set was short, strong year classes were identified and were extremely useful for modelling purposes. There was some question as to the validation of rings in tailor as the Western Australians age the same otoliths differently toscientists within Queensland. Bream and flatheadoto liths can be fairly easily aged, but there was much difficultywith the whiting species.

The resolution of the logbook database was often a major stumbling block in developing stock assessment models. Much of the data is still entered within the 30-minute square grids, whereas many management rules require at least 6-minute square or direct latitudes and longitudes. Furthermore,in many cases the data can not be resolved to species level. The most chronic of the species investigated was the inshore whiting. Gear type beyond net and trawl etc. is alsonot recorded even though, forexample, in the net fishery, beach seine gear have extremely different catchability and selectivity functions compared to tunnel and gill nets.

Modelling techniques used within the workshop were extremely varied, from simple biomass dynamic models, estimation of total mortality fromage information, generalised linear models to a full-scale relative age model using survey, tagging and catch rate data to estimate population size and recruitment indices. The workshop had limited success in estimating final population parameters to determine the health of the stocks. Mostly this was due to logbook problems (e.g. resolving whiting catches into species), lack of contrast in the data (e.g. easternking prawns catch rate trends), defining effort(e.g. the multi-species and multi-endorsed net fishery) or unknown recreational catch (e.g. tailor and the inshore finfish species). However, of extreme use were the management questions that were addressed. The benefitsof a winter closure forscallops, the benefitsof opening for a short period the scallop preservation zones to catch senescent animals and testing the spanner crab TACC decision rules are examples. It can be argued, however, that the three greatest benefitsof this workshop were: • the collation of data, which prior to the workshop was either untraceable, on old computers, only on datasheets or in various filesheld by the scientists involved on the project at the time, and • the communication between all interested individuals, be they biologists, stock assessment specialists, fishers etc., concentrating on species of relevance to south east Queensland in a hands on forum, and • a list of research priorities and monitoring directions for the future.

11 INTRODUCTION

Figure 1.1. Ms Cathy Dichmont (workshop convener) and Dr Malcolm Haddon (workshop facilitator)

1 A STOCK ASSESSMENT REVIEW resources at levels commensurate WORKSHOP was held from the 16-28 with the value of the resources and August at the Southern Fisheries Centre, risk of overfishing. Deception Bay, to investigate possible stock assessment techniques that could The focus of the workshop was therefore be applied to available fisheriesdata. upon the stock assessment of each The workshop was funded by the species and how this could be used to Fisheries Research and Development improve on management advice. Corporation (FRDC), Departmentof The resources chosen for review were Primary Industry, Queensland (DPI) determined through a series of Resource and the Queensland Fisheries Manage Priority workshops held within DPI. ment Authority (QFMA). It was Those workshops concluded that the convened by stock assessment scientist primary southeast Queensland resources Cathy Dichmont of DPI, Southern are spannercrabs (Ranina ranina), Fisheries Centre and facilitated by easternking prawns (Penaeus plebejus), Malcolm Haddon from the Australian saucer scallops (Amusium japonicum Maritime Collegewho is an expert in balloti), sea mullet (Mugil cephalops), this field. Participants outside the tailor (Pomatomus saltatrix) and an Centre were invited fromthe fishing inshore finfishcomponent of bream industry, SUNFISH, Queenslands (Acanthopagrus spp), whiting (Siliago CommercialFishermen's Organisation spp.) (excluding the trawled stout (QCFO), Queensland FisheriesManage whiting) and dusky flathead (Platy ment Authority (QFMA), DPI Northern cephalus fuscus). A finalopen-day series Fisheries Centre, New South Wales of seminars was held on 28 August Fisheries, NorthernTerritory Fisheries, reporting on the results of the workshop. WesternAustralia Fisheries and CSIRO All sessions were well attended in Cleveland and Hobart (see complete especially the final seminar series. list of participants and their affiliation in Chapter 11 ). The same formatfor proceedings was The workshop had fourmajor used for each species or species group objectives: examined. The first step was an intro duction to the biology of the species, • To review and evaluate existing its current management arrangements, stock assessment and monitoring the data available and the research that programs and methods relating to had already been conducted. Following fished resources in SE Queensland. this introduction, fouror five tasks were • Use existing data to develop and test proposed which smaller sub-groups alternative assessment techniques. could most likely pursue profitably. The members of the workshop would then be • Makerecommendations on future sub-divided among these activities and directions and/or priorities for each such sub-group would work on research and monitoring, and, where their particular problem forbetween possible, to advise the Queensland four and six hours. Each sub-group Fisheries Management Authority would then report back to the complete andDPI Fisheries Group of the workshop. On the basis of these reports biological status of the resources and the ensuing discussion, the final to be investigated. recommendations and suggestions of the • To develop appropriate strategies workshop would be devised and agreed for monitoring and assessment of upon. The workshop report followsthe southernQueensland fisheries same structure.

2 The above approach was surprisingly ment. Ifyou look at what has happened successful. Efforts were made to ensure over the last 20 yearsor so, it has been a thateveryone would agree that the really interesting process in findingout recommendations produced were a about our fishstocks. I know when I was consensus of opinion. Many of the doing research, it was really exciting to sub-groups working on particular find out about how these things migrate problems managed to produce useful and I remember the first time people results although time limitations within could pin up a life cycle of king prawns. the workshop meant that none fully We tried to understand the biology of completed their tasks. The most these things. promising of their analyses will be pursued further andresults reported at a We've come past that now and we have later date. The aim of this current report got a pretty good handle of the basic is to describe the workshop and its biology of most of these animals, but activities. Given the time available, we are now being asked, through the all results in this document are of a management process particularly, what preliminary nature. thestatus of these stocks are for example arethey healthy or not so healthy. That is the challenge I guess for the scientists here. It is quite important to come to 1.1. WORKSHOP grip with this challenge. INTRODUCTION Another thing developed over these Below is the transcript of the Workshop last few years is that we now have a opening speech by Dr BarryPollock, reasonably good commercial logbook DP/ Fisheries General Manager. program in Queensland. We understand that there are some problems and that it It is really a privilege to have a work is not perfect, but at least we've got one shop like this at the Southern Fisheries we can work on and try to improve. Centre. This workshop is basically being funded by the Fisheries Research and More recently, we've started to get a Development Corporation (FRDC) database on recreational fishing, but it and we really appreciate that funding. has only been going for a year. We Queensland Fisheries Management really need to get catch and effort Authority (QFMA) and Queensland statistics on that sector. Also, I guess, Department of Primary Industries (DPI) information fromthe charter sector and have also been supportive financially, smallersectors. We have to get the but I guess the participants will make it complete picture. work. The workshop is going to look at stocks in south Queensland on spanner What we are not doing and therefore it is crabs, eastern king prawns, scallops, the challenge, is a more sophisticated or mullet, tailorand a finfishassemblage a more elaborate monitoringprogram. of bream/whiting and flathead.That is a We don't have anything now. We've big program you have given yourself done some research in thelast few years and it will be interesting to see what on stock like the finfish, scallops, barra comes out of it. mundi, mud crabs and so on. FRDC has fundeda lot of those short term, 3 year We areat a really interesting stage in research projects. We need to make Queensland with our fishstocks, in that some important decisions now as to at the moment the emphasis is for what we are going to do especially accurate more rigorous stock assess- in terms of longer-term monitoring.

3 It has been quite exciting over the last two or 3 years. It is a good start but the couple of months. We have come up amount of modelling in the book is with a concept to get up a monitoring pretty minimal and that should be team, which will be a dedicated group corrected in time. of people who will monitor various indicators of key fish stocks. These I wish the workshop well. There will be indicators may be fisheryindependent some real implications forresourcing, or it may be fishery dependent. We especially the monitoringside and it will haven't really worked out the details give us a picture of what we know about and that is why this workshop is these really important stocks at this point important, because it will give us some in time and how good the data is. clues as to what andhow we might actually monitor. 1.2. AUTHORSHIP NOTE We don't have a lot of resources, but at least it is a start and I am hoping that Numerous people have contributed to eventually we can entice some money these proceedings. Unless otherwise from industry. We hope that this team stated below, the authors areCathy would startin June of next year, so we Dichmont and Malcolm Haddon, have a yearto work out exactly what we whereas the Proceedings was edited need to do. by Cathy Dichmont, Malcolm Haddon, Kate Yeomans and Kath Kelly. The other thing that interests me about this type of work, is how we commun CHAPTER 2 SPANNER CRAB icate the results to the fishing industry This chapter was written in collab and the public. We are dealing with a oration with Ian Brown. Section 2.2 public resource. Insome studies, science "Introduction" was written by Ian Brown that is quite complex have had a fan and is a summary of workshop present tastic extension program. For example, ations given by Ian Brown, Cathy scientists have built up a visual model of Dichmont, John Kirkwood, Richard Moreton Bay andBrisbane River and Freeman and Mark Doohan. tried to build the complex science into a simple model which is visually strong. CHAPTER 3 EASTERN KINGPRAWNS This chapter was written in collaboration Interms of extending our assessment with Tony Courtney. Sections 3.2.1 work, it is maybe not your role, but it "Biology", 3.2.2. "Description of certainly is something into which you fishery", 3.2.2.a. "Introduction" were should have input. How do we get written by Tony Courtney on which he informationout to people who we may based his presentation given at the want to convince? I don't think we can workshop. Section 3.2.2.b "The just stop at the scientific, we need Development of the Fishery" was extension beforepeople really accept written by Peter Gaddes and was a tough message. presented by Peter Gaddes and Peter The last thingI want to mention, is Seib and 3.2.2.c "Management" was that we have already produced the written and presented by Mike Dredge. 'Condition and Trend' document which CHAPTER 4 SAUCER SCALLOP is not aimed at the scientific community, but at a broader community and fishers This chapter was written in collaboration particularly. We are committed to with Mike Dredge. Section 4.2. producing something like this every "Introduction"was writtenby Cathy

4 Dichmont and Mike Dredge and is a CHAPTER 6 TAILOR summary of presentations given by This chapter was written in collab Cathy Dichmont, Mike Dredge and oration with Simon Hoyle. Section 6.2 Richard Gilbert. "futroduction" is a summary of presentations by Simon Hoyle, Ian CHAPTER 5 SEA MULLET Halliday and Michael O'Niell and was This chapter was written in collab written by Ian Brown and Simon Hoyle. oration with IanHalliday. Section 5.2 "futroduction" is an extract fromthe CHAPTER 7 BREAM, WHITING, Draft Final Report to FRDCf or Project FLATHEAD No. 94/024 'Assessment of the Stocks This chapter was written in collab of Sea Mullet in New South Wales and oration with Simon Hoyle. Section 7.2 Queensland Waters'. The Queensland "futroduction"is a summary of section was writtenby Ian Halliday and presentations by Simon Hoyle, Ian Kath Kelly and the NSW information Brown, DarylMcPhee and Darren was written by John Virgona, Kerrie Cameron and written by Ian Brown Deguara and Darryl Sullings, NSW and Simon Hoyle. Fisheries. The workshop presentation by Ian Halliday was based on this report. CHAPTER 9 AVAILABLEDATA This chapter was compiled andwritten by Kate Yeomans.

5 6 common name SPANNER CRAB

Ranina ranina

7 2.1. PARTICIPANTS

Dr Ian Brown (Fisheries Scientist, DPI, Deception Bay) Mr Rik Buckworth (Fisheries Scientist, DPIF, Darwin) Dr Tony Courtney (Fisheries Scientist, DPI, Deception Bay) Mr Mark Doohan (Crab Resource Manager, QFMA,Brisb ane) Ms Cathy Dichmont (Fisheries Scientist, DPI, Deception Bay) Dr David Die (Fisheries Scientist, CSIRO, Cleveland) Mr Russell Doyle (Spanner crabber and Crab SAG/MAC representative) Mr Mike Dredge (Industry Manager,DPI, Deception Bay) Mr Richard Freeman (Spanner crabber and Crab SAG/MAC representative) Mr Dale Golian (Fisheries Manager, NSW Fisheries, Sydney) Dr Neil Gribble (Fisheries Scientist, DPI, Cairns) Dr Malcolm Haddon (Head, Department of Fisheries, AMC, Launceston) Dr Burke Hill (Fisheries Scientist, CSIRO, Cleveland) Mr Simon Hoyle (Fisheries Scientist, DPI, Deception Bay) Ms Kath Kelly (FisheriesScientist, DPI, Deception Bay) Dr Steve Kennelly (Fisheries Scientist, NSW Fisheries, Sydney) Dr John Kirkwood (Fisheriesscientist, DPI, Deception Bay) Dr David Mayer (Biometrician, Animal Research Institute, DPI, Yeerongpilly) Mr Dave Mitchell (QFMA Crab MAC Chair) Mr Michael O'Neill (Fisheries Scientist, DPI, Deception Bay) Dr Jenny Ovenden (FisheriesGeneticist, DPI, Deception Bay) Dr Andre Punt (Fisheries Scientist, CSIRO, Hobart) Dr James Scandol (Fisheries Scientist, University of Sydney as spanner crab stock assessment representative of NSW Fisheries, Sydney) Mr Clive Turnbull (Fisheries Scientist, DPI, Cairns) Ms Kate Yeomans (Fisheries Scientist, DPI, Deception Bay)

8 2.2. INTRODUCTION boat, and was therefore very attractive to potential crabbers. Spanner crabs (Ranina ranina) are found throughout the tropical Indo The first significant commercial Pacific region fromthe easterncoast landings were made early 1979. of South Africa through the Indonesian However, it is difficultto establish Archipelago to Japan and Hawaii. In precisely how the fisherybegan and who Australia, the species occurs fromthe was involved, as a number of operators Abrolhos Islands (WA) throughnorthern lay claim to having been the first to tropical and Great Barrier Reef waters fish spanner crabs commercially. A and as far south as southernNSW. They considerable proportion of the initial have been fished commercially around landings is attributable to unlicensed the Hawaiian Islands since beforethe fishermen,particularly in the Second World War. Catch statistics Mooloolaba-Caloundra area where the compiled by the Hawaiian Division of fishinggrounds were relatively close to Fish and Game indicated an average the coast. As a result, early 'official' annual production before the mid-1980s catch statistics, maintained by the of approximately 10 tonnes, with a peak Queensland Fish Board,are highly likely of 35 tonnes in 1972. The species is also to have underestimated the actual exploited to a minor extent around the landings. The total annualcatch of Philippines and the southerncoast of spannercrabs in Queensland during Japan (northwest PacificOcean) and in 1983-84 was estimated to be around 300 the Seychelles Islands (westernIndian tonnes. This is an order of magnitude Ocean), but production figuresfor these greater than the highest annual catch regions are unavailable. A largespanner reported fromthe Hawaiian fishery crab fishery which has recently during its 35 year post-war history. The developed in Thailand, is a significant construction of the fishinggear used had competitor to the Australian fishery. by thistime changedfrom the conical dilly to a very much flatterversion, with The fishery forspanner crabs in the mesh stretched relatively tightly Australia has a relatively recent history, across the frame. This development going back only two decades. It began resulted fromthe fact that the crabs to develop as a commercially viable became so entangled in the loose mesh operation around 1978-79 when it was o that it was very difficultto extricate f und that crab 'dillies', previously used them without severely damaging the net. by recreational fishersto catch mud crabs (Scyllaserrata) and sand crabs With the influxinto the fishery of more (Portunus pelagicus) in estuarine professional crabbers, who generally habitats, were a very effective means marketed theircatch through the Board's of catching spanner crabs in offshore regional depots, 'official' figures by the waters. In the early days of the fishery, mid 1980s accounted fora more realistic inverted dillies or 'witches hats' were proportion of the State catch. At that used exclusively. This apparatus stage the recreationalcatch was believed comprised a baited, conical, mesh net to be small butstill significant, and attached to a circular metal frameand professionalcrabbers were starting to supported at the apex by a small net develop alternative outlets for their float. Nets were set on the sea floor catch, including cooperatives, processors individually, each with a separate buoy anddealers. line and surface marker float. The equipment was simple to construct, Inthe mid 1980s fishersin northern inexpensive, easy to deploy froma small New South Wales began to realise the

9 potential of the spanner crab resource, Fisheries Centre conducted a 3 day and significantlandings began to be survey in the northernHervey Bay area reported by several of the Northern in early November 1987. The survey Rivers fishermen'scooperatives. During revealed the presence of a potentially the s;ime period the Queensland fishery fishable stock of Ranina raninain the expanded south to the border and north area between Bundaberg, Lady Elliott to Double Island Point. The results of a Island, and Round Hill Head (Seventeen voluntary logbook program, run during Seventy). This informationwas made 1982-84 and 1987, showed an increase available to a number of crabbers during in individual effort and indicated a the subsequent two or three years, and decline in catch rates. The average doubtless contributed to the northward number of net lifts per boat per day expansion of the fishery. increased from about 90 in 1982-83 to 198 in 1987, while the average In the late 1980s there was a major catch rate, after an initial increase from advance in the collection of fishery 3.5 to 4.5 legal sized crabs/net between statistics in this State with the develop 1982 and 1983, declined steadily to ment of the compulsory CFISH logbook about 1.5 crabs/net in 1987. Legal sized program by QFMA and DPI. This crabs have a minimum carapacelength program is now incorporated into the of 100 mm. This CPUE would translate Queensland Fisheries Information to approximately 0.6 kg/net lift. This System - QFISH. Although the level decline in CPUE was thought to have of detail was less than had been obtained been due to a real decrease in stock by the voluntary spanner crab logbook, density (particularly in the geograph there was at last a mechanism available ically 'central' part of the fishery that for estimating total State catch with had been subjected to exploitation for some confidence. the longest time). In response to concernsabout thefuture The fisheryhas always been considered sustainability of the fishery,a warning a convenient 'alternative' fisheryto about further investment in the fishery most of the active participants, many was issued in January 1994. The fishery of whom hold net, line and crab licence had faced an unprecedented increase in endorse-ments. This means that under both the number of boats entering the present regulations there is enormous fisheryand the total number of fishing potential for boats to move in and out days, with a resultant increase in overall of the fisheryat will (e.g. from thesand landings. In December 1995, two crab, mud crab and reef line fisheries). Management Areas were declared. As early as 1988 there was general Management Area A (South of 23S agreement among those involved in and east of 151.45E) was generally the management of the spanner crab accepted as the developed fisheryand fishery, that the major management was subject to a Total Allowable Catch problem in the fisherywas the lack of (TAC) and a maximum daily catch a practical way to reduce or even limit limit of 300kg. Several changes to the the number of actively participating daily catch quota have taken place and vessels in the fishery. the fishery in this region is moving to an Individual Transferable Quota (ITQ) In response to requests by interested managed fishery. Management Area B fishermenfor information on the (South of 23S and west of 151.45; north northerndistribution of spanner crab of 23S) is basically unrestricted as it is stocks, research staffat DPis Southern seen as a developing fishery.Closed

10 seasons have been imposed over time, 2.3. SPANNER CRAB the most important of which is the WORKSHOP TASKS spawning closure between 20 December and 5 January. Five problems or issues were investigated by separate groups: Attempts at production and size based 1. Effort standardisation. modelling have not been immediately successful. This is mainly due to the 2. Develop a YPR model to investigate unstandardisednature of the effort evidence of growth overfishing. as well as lack of definitive growth 3. Investigate the possibility of deriving information. Some growth analysis a TAC froma review of the spatial of the species in the Seychelles Islands development of the fisheryand the have been attempted, but only with a expected yield per area. very small data set. Analysis from a tagging study completed in New South 4. Develop a simulation model to test Wales indicated that spanner crabs might the robustness of the proposed be much longer lived than previously Decision Rules. thought. However, little informationis 5. Examine all existing growth and availableon earlygrowth. A recent mortality data and synthesise a study by DPI with dredges has had growth curve. success in samplinganimals smaller than 50 mm carapace length, which are 2.3.1. EFFORT STANDARDISATION assumed to be less than two years old. Information froma questionnaire on None of this work had been combined vessel characteristics, gained from a into a single model of spanner crab section of the fishingfleet at the end of growth. 1997, is now availablefor analysis together with thecatch and effort data in Management Area A was subdivided CFISH. A OLM approach was applied into fiveregions that emulate the using the statistical package GENSTAT. expansionof the fishery. Due to a Some preliminary cleaning of thedata lack of an objective stock assessment was needed to remove suspect effort and method, which would guide the setting catch records fromconsideration. The of the annual TAC, a 'ratio-metric' dependent response variate was ln (catch model was used to estimate the TAC. rate), where catch rates were kg of It assumed that the constantcatch rates spanner per net lift. Catch rates were over the last fewyears in Areas 3,4 and log-transformed as should be done with 5 could be interpreted as indicative of this type of analysis but also because the sustainable catches. These catches per data clearlywere lognormally unit area were multiplied to the whole distributed. of Management Area A region. A set of decision rules have been developed Catch rates are often assumed to be by the Crab SAG in order to set the proportional to stock size. In a OLM, the future annual TAC in a manner year factor is treated as a surrogate for transparent to everyone. This process stock size. Unfortunately, manyfactors may become part of the Draft Crab other thanstock size can influence Management Planof QFMA. catch rates. The objective of effort standardisation is to account forthe variation in catch rates introduced by factors such as 'skipper experience', 'vessel size', 'fishingregion', etc. In this

11 preliminary analysis thefactors tested 2.3.2.a Depletion Event. for influence upon catch rates were: Discussions in the workshop led to year, management region, crew number, industry members remembering a skipper experience, hull length and hull particular event where many fishers cruising speed. The data on navigational converged on a single area andfished it aides obtained fromthe questionnaire hard until catch rates declined markedly. lacked contrast and had many missing It was considered that, if the data from records so these factors were not this 'event' could be foundin the included in the analysis. Region, year database, then it might be possible to and their interaction term dominated the conduct some formof DeLury depletion analysis outcome. Skipper experience analysis on the decline in catch rate was a significantfactor, with the through time. The database was adjusted mean catch rate amongst investigated and the 'hot spot' was f 'experienced' fishers (0.96 kg.lif 1) discovered to be near North West Island being substantially higher than that of reef. This area was fishedvery heavily the 'novice' group (0.71 kg.liff1). Hull by the commercial fleet over an eight length may be a significantfactor, but week period during October-November time prevented an in-depth analysis of 1996. During that time, some 40 tonnes this factorand of the effectsof crew of spanner crabs were taken out of an number and hull speed. area approximating a 6 minute grid square. The appropriate catch and This preliminary analysis has been effort data were retrieved and fitted to a extended and completed by Ian Brown standard Leslie depletion model (Figure and David Mayer. Results have been 2.1), which provided a catchability reported on in detail in FRDCReport coefficient and an estimate of biomass 95/002 (Brown, I.W. et al., 1999). in the depletion area. The regression parameterestimates are given in Table 2.1. 2.3.2. DEVELOP A YIELD The negative of the slope of a linear PER-RECRUIT MODELTO regression of the graph provided a INVESTIGATE EVIDENCE catchability coefficient of 0.000034. OF GROWTH OVERFISHING This means that one unit of effort (a This group completed two sub-tasks; an dilly lift) captures, on average, 0.003% investigation of a depletion event anda of available biomass in this particular YPRanalysis. aggregation of spanner crabs. Whether this figure is typical of all Queensland spanner crab fishing grounds needs furtherinvestigation.

Table 2.1. Linear regressionparameter estimates used in theLeslie analysis of the commercial depletion offNorth West Island reef in Management Area B Parameter Estimate Std. Error P-va/ue Lower95% Upper95% Intercept 1.736 0.142168 6.75E-15 1.448063 2.023187 Slope (-q) -0.000034 0.000005 8.58E-08 -0.000044 -0.000023

12 The estimate of total biomass available immigration into the depletion area in the depletion area is calculated by would falsely increase the estimates of extrapolating the regression line to the the total biomass and decrease the point where it would cross the x-axis estimate of the catchability coefficient. (total catch). In this case, an estimated pre-exploitation biomass of 58 tonnes 2.3.2.b Yield-Per-Recruit was calculated. It should be remembered The second task of this group was to that thiswas considered an exceptional develop a standardequilibrium-based aggregation of spannercrabs and so YPR model of the resource. Input might best be viewed as an estimate of parameterswere: the upper limit of crab biomass. 1 • ) natural mortality1 (M year- of The Leslie depletion analysis assumes 0.38 year- based on the results of that the average catchability remains the group reported in Section 2.3.5, constant for the duration of the depletion • age at recruitment to the fishery event. It also assumes that no (tr) of 1.5 years, immigration or emigration of spanner • crabs occurred during the fishdown. weight at L (W...,) of 1.5 kg and This analysis may therefore have a • intrinsic growth rate, K, of 0.28 year-1 positive bias because of the methodof (fromgroup results reported in capture uses bait so the area of influence Section 2.3.5) may expand through time. Any

3.5

3.0

y = 1. 7356 - 0.00003x - 2.5 2 ·- • R = 0.5248 - 2.0 0 1.5 • • • 1.0 • • 0.5 •••• 0 0 10000 20000 30000 40000 Cumulative Total Catch (kg)

Figure 2.1. Commercial depletion of a newly discovered area offNorth West Island reef in Management Area B between 1 October and 20 November 1996 using CFISH logbook data.

13 5.2

3.6 Age at first Capture tc

0 FishingMortality

Figure 2.2. Plot of yield-per-recruit of spanner crabs for a range offishing mortalities 1 (year- ) and age-atjirstcapture (t) in years.

A three-dimensional plot of YPR for present uncertainties associated with variouscombinations of age-at-first estimated growth parameters and natural capture (from0.4 to 7 years old) and mortality rate. 1 fishing mortality (from0.2 to 3.0 year- ) is in Figure 2.2. Indicated maximum 2.3.3. INVESTIGATION OF YPR occurred at anage of firstcapture THE APPLICATION OF A of about 2.4 years over a wide range of PRODUCTION/YIELDMODEL fishingmortality rates (F) from 1.2 to 1 This group's objective was to determine 3.0 year· • This is consistent with our whether a defensible TAC could be understanding of the age of a spanner estimated. The approach used was to crab at the minimum legal size. Itcan consider the spatial development of the therefore be concluded that growth fisheryand findthose areas fishedthe overfishing is not likely to be a potential longest. If the level of yield takenfrom problem in this fishery, subject to those areas had remained relatively

14 constant through time one canassume The work attempted by this group was that such yields could continue to be ambitious in its scope andthey ran out taken on an on-going basis i.e. the of time. To complete the work, the present yields are sustainable. If there is frequency distribution of daily catch no evidence of serial depletion, then the rates in particular space units should be average yield per spatial unit may determined. A highly skewed lognormal provide an estimate of TAC. distribution would indicate depletion over time. Also, if this distribution were The ESRI GIS package ARCVIEW was compared fromone year to the next, used to plot the yearly catch andeffort then a truncation of the upper catch rate by 30 minute squaregrid. These clearly levels would again indicate depletion. demonstrated the expansion of the However, the degree of this depletion fishery and the establishment of key may not be established. ports (e.g. Figure 2.3 a and band 2.4a and b).

15 Spanner Crab Catch (kg's) 100 - 10000 10000 - 50000 50000 - 100000 100000 - 200000 200000 - 400000 400000 - 600000

1988catch 1989catch

19rocatch 1991 catch

Figure 2.3 a. Spanner crab catch (kg) distribution over the Queensland coast for years 1988 to 1991 using the raw data from CFISH.

16 1992 Catch

Figure 2.3b. Spanner crab catch over the Queensland coast for years 1992 to 1997 using the raw data from CFISH. Legend as per Figure 2.3a.

17 Spanner Crab Effort (lifts) 100 1000 1000 - 10000 10000 - 50000 50000 -100000 1111100000 - 200000 1111200000 - 900000

LJ 1988 Effort 1989 Effort

1990 Effort 1991 Effort

Figure 2.4 a. Spanner crab effort (lift) distribution over the Queensland coast from 1988 to 1991 using the raw datafrom CFISH.

18 1992 Effort

1995 Effort

Figure 2.4 b. Spanner crab effort (lifts)distribution over the Queensland coast for years 1992 to 1997 using the raw data from CFlSH. Legend as per Figure 2.4a.

19 2.3.4. DEVELOP A SIMULATION The decision rules were vigorously MODEL TO TEST THEROBUSTNESS tested using a model with 21 input OF A SET OF DECISION RULES parameters (Table 2.2). This group had the objective of The strategy used to test the decision developing a strategyand algorithmfor rules was to develop an operational testing the operational characteristics of model describing a hypothetical sets of decision rules, such as foundin population which was effectively Fishery Management Plans. A set of sampled to provide catch and effort data. decisions rules, as developed by the The management decision rules were Crab Stock Assessment Group, were applied to the informationderiving from tested with a number of sources of the operational model and the resulting uncertainty and error: global TAC was fed back into the 1. highly variable recruitment; operational model foranother year's 2. inaccurate estimates of catch rates activity (Figure 2.5). The rules were (i.e. large observation errors); tested over a 25 yearperiod.

3. inappropriatemanagement regions 2.3.4.a OperationalModel with respect to the actual stock structure; The system was modelled to have four management areas and four stocks. 4. a model assumption was that catch The management areas did not align, rate was proportional to biomass, however, with the biological stocks: whereas, in reality, catch rates were stocks 1 and 2 were regarded as being only proportional to biomass at low a single management area; stock 3 was density - Theythen remained regardedas two separate management relatively constant at increasing areas, and stock four was correctly densities (i.e. hyperstable). identified as a single management area.

Operation Model Catch & Effort Management Model �... - Stock structure by Management - Decision rules to - Recruitment Area set TAC - Stock dynamics - Catch by stock �.... Global TAC

Figure 2.5. Schematic describing the relationship between the perational model used to . � define the hypothetical population and the management regime imposed.

20 A first order difference model was used Since only a global TAC is set by the to describe the dynamics of the resource decision rules, within the operational from one year to the next: model a rule was appliedto divide the catch taken from each stock: Rand , e cr aJ1 1 ].s and 1 B e' cru TAC c ,s + ( B t,s I :2.4 ""B randcru [3 e + ;; £.i /,S :2.1 where T AC is the Total Allowable where B ,s is the biomass at time, t, t t Catch that was set foryear t, of stock, s, O'u is the variance of fisher p is a proportional constant, knowledge. Mis natural mortality B ,s is the biomass at time t C ,s is thecatch taken from t t for the biological stock, s. stock s, at time t, Randis a random normal This implies that the fishery takes deviate (Z), the most yield fromthe areas with the O'r is recruitment variance, highest levels of biomass (despite having a is a stock-recruit parameter, incomplete knowledge of the biomass p is a stock-recruit parameter level). The initial stock biomass in each and calculatedby: area was definedas an input parameter (Table 2.2). (1- 81).2 = : 2.2 [3 (8 0.2) 2.3.4.b Management Model All the decision rules that could where o is the steepness of the potentially alter the TAC were driven by recruitment relationship catch rate trends. Three of the selected Ks is the stock carrying capacity. decision rules recommended that the TAC should be reviewed. As a direct Catch rates of each management area decision was required by the model, were defined as: deterministic decisions about TAC levels were defined. Consequently, if one of these three decision rules was triggered, a change in TAC was agreed where Bt,z is the biomass at time t upon. In reality, this may not always be forthe management zone, z, the case. yis the index of hyperstability, randis a random normal 2.3.4.c Testing the Decision Rules deviate (Z) Several scenarios were tested. In one O'q is the variance of the case, a seriously depleted resource was catchability coefficient. simulated. In this case, the decision rules performed extremely well. They A small number (0.00001) was added did not allow the stocks to go extinct and to equation 2.3 to avoid taking the reversed the downward trend of natural log of zero. The effects of population biomass. different degrees of hyperstability (y) were tested. A second scenario described a resource in which one of the fourstocks was

21 severely depleted and/or had a much An example of a single run is given in smaller carrying capacity with respect Table 2.2 below. The resultant catch rate, to the others. The decision rules saved stock recruit relationship, annual TAC the other stocks, but were unable to and biomass trendfor each stock are prevent the extinction of the seriously given in Figure 2.6. depleted stock. Table 2.2. Parameter values for an A further extreme case of the resource example run testing theSpanner Crab being extremely large and the TAC Draft Management Plan TAC setting being well below sustainable levels were decision rules. tested. In this case the decision rule did Parameter Value Parameter Value not respond well at all and were unable p 1.4 Kl 10 000 to take advantage of the good resource M 0.38 K2 10 000 biomass and production characteristics. y 0.5 K3 10 000 B 0.8 K4 10 000 The simulation model indicated that a 0.04 R1.1 425.9 the set of catch rate based decision rules 5 which were tested were very conserv f3 0.06 R12 425.9 ative, precautionary and risk-averse. 7 They are likely to perform extremely Bl,l 100 R, 425.9 well in a declining fishery, but may Bl,2 400 R,,, 425.9 underutilise the resource when it is in Bl,3 400 cr- 1 good condition. Furthermodelling of Bl,4 5 crr 0.2 this type will be needed to do a more 000 in-depth analysis of these rules, but some cr" 0.2 modificationof these rules is needed.

22 14000 12000 810000 (I) (I) 8000 .2 6000 4000 2000 0 0 5 10 15 20 25 No. Years

a: 1200 1000 0 a, 800 .. 600 :ii: 400 200 s0 I- 0 0 5 10 15 20 25 No. Years

160 w 140 120 0 100 ai 80 .DI 60 a:a, 40 20 0 0 5 10 15 20 25 No. Years

8000 _ 7000- :;:::" 6000 i 5000 .54000 ·2 3000 i 2000 1000 0 0 5 10 15 20 25 No. Years

Figure 2.6. Stock biomass, annual TAC, regional catch rates and stock recruitment for a single example run that test the hypothetical set of Decision R111Ps. 23 2.3.4.d Conclusions tagging data (which only includes crabs The strategy of using an operational about 65 mm carapace length and larger) model to interact with a selected with the Queensland dredge data (which management procedure/plan was easy samples animals smaller than 60 mm in to implement, simple to manipulate and carapace length). A conversion factor the outcomes were straightforward to (John Kirkwood personal interpret. This approach provides a communication) was used to convert relatively simple mechanism for orbital carapace length, used in NSW, conducting a heuristic searchfor an to rostral carapacelength, used in optimal set of decision rules. Such Queensland. The dredge data, arguably, an heuristic search was not carriedout shows two distinct cohorts (Figure 2.7). but should be done fora true lifeFishery There was consensus that they were not Management Plan. two or more years apart, as there are no intermediate values. One cohort is therefore assumed to be less than 1 year 2.3.5. EXAMINEALL EXISTING old (O+ old) and the other between 1 and GROWTH AND MORTALITY DATA 2 years of age (1+ old). Each analysisof growth of spanner crabs throughout the world and in Australia The only available adult data for uses differentdata and has produced Australia is froma tagging study very differentgrowth curves. No completed in New South Wales. This complete growth curve is available clearly showed that males and females foruse in Australia. Informationabout have different growth rates. Three growth can affect one's view of the differentplots of mean growth from the animal's longevity, which, in turn, affects tagging model developed by Chen and one's view of its ability to sustain a Kennelly (1999) on the NSW tagging certain level of exploitation. This group data, superimposed with the dredge data tried to finda way of linking the NSW were attempted withall combinations

a, .0 20 E :::, C 15 � C a, :::, 10 a,

0 5 10 15 20 25 30 35 40 45 50 55 60 Carapace Length (mm) Figure2. 7. Size-frequency distribution ofjuvenile spanner crabs caught in a Queensland dredge.

24 fromthe tagged animals being 1 + year 0.270, and a to= -0.21 (a lead time of 2 old to 4+ year old were considered years gives L"" = 142.2, a K = 0.279, and (Table 2.3). a to=-0.20). It is clear that the sum of squared Ideally, each sex should be treated residuals were minimised when the separately. However, if one assumes tagged animals were assumed to start as that the averaged curve gives a 2+ animals for each gender (Figure 2.8). meaningful representation of average When a search is made for the exact age size at age, natural mortality is likely to 1 increment (between the oldest dredge be in the range of 0.31to 0.46 yea{ , animals and the youngest tagged equating to a maximum age of 15 to 10 animals) which produces the smallest years of age. These analyses are sum of squared residuals, approximately preliminaryand require more data on 2.4 years results for both males and juvenile growth rates and more females. information on the moulting frequency of the adults. The average von Bertalanffy curve for both sexes combined, using a lead age of 2.4 years, leads to an L = 141.1, a K =

Differentassumptions concerningthe start age for youngest tagged animals Males Females 1+ 2+ 3+ 4+ 1+ 2+ 3+ 4+ L 170.3 168.5 166.0 162.4 119.9 117.6 114.8 111.7 K 0.258 0.229 0.205 0.190 0.348 0.334 0.324 0.324 To -0.161 -0.201 -0.248 -0.294 -0.188 -0.206 -0.226 -0.237 SSQ 3100.6 1504.0 1561.5 2190.6 2166.3 1311.2 1316.3 1582.4

Table 2.3. Von Bertalanffy growthparameter estimates of Queensland dredge and NSW tagging data given differentassumptions of start age foryoungest tagged animals.All curves were obtained by including the dredge samples as unsexed juveniles.

25 180 160 140 120 -"'O ___ Q •• E - - - _ - 100 - - - 'tl CJ) 80 • Tagged males o Tagged females ..J 60 x Dredge juveniles 40 -- Female Model 20 ---Male Model 0 0 1 2 3 4 5 6 7 8 9 10 11 Age (years)

Figure2.8. Spanner crab growth data from two combined sources (Queensland dredge and NSW tagging data) for males and females.

2.4. CONCLUSIONS A small-scale depletion event was discovered in the database and a Leslie depletion analysis estimated a pre GROUP REPORTS 2.4.1. exploitation biomass over the areaat the Much of the discussion centred on time. There was agreement that this the possibility that catch rate is not result could not be extrapolated over the proportional to biomass in this fishery. whole fisheryarea, but could be seen as Preliminary analysis at the workshop an example of high biomass areas. using a generalised linear modelling Yield-per-recruit analysis demonstrated approach showed that effort standard that growth overfishingis not likely to isation of regional and temporal be problem in the spanner crab fishery, changes, vessel characteristics and level subject to uncertainties associated with of skipper experience could explain estimated growth parametersand natural about 27% of the variance. There was mortality rate. general agreement that animal behaviour in relation to the baited dilly and oceano An analysis of the spatial distribution graphic factors would greatly influence of catches through time highlighted the catch rates. Furthermore, the move to expansion of the fishery and supported an ITQ system could affectthe temporal the approach in which different"assess integrity of the catch-effortdata series. ment regions" within Management Area A fishery-independent survey was seen "A" are analysed separately. as essential, but would not be entirely without the problems above unless an The spanner crab T ACC decision rules active sampling method is used to catch were tested using a model population, spanner crabs. which was harvested and then managed in accordance with the decision rules.

26 To provide a stringenttest the model These doubts stemmed from the included high variability in recruitment complex interactions between the and catch rates errors, incorrect assump crab behaviour relative to the baited tions concerning stockstructure relative dillies, between fishersand crab to management regions and non aggregations, and the impact of proportionality of catch rate relative changes of current speed and to biomass. This testing strategy direction. The workshop strongly demonstrated that the decision rules recommended that independent perform well under adverse conditions surveys of spanner crab biomass to protect the regions fromcollapse, but were needed forlong term tended to underexploit recovering or monitoring in order to provide an healthy regions. alternative, more defensible index of abundance. An attempt was made to combine the Queensland dredge samples of juveniles b. The institution of an ITQ with the NSW tagging data of adults into management system is likely to a single von Bertalanffy growth cause further problems with using function. Based on an assumption that and interpreting the catch rates of two clear modes in the dredge data are spanner crabs forassessment from animals thatare o+ and 1 + old, purposes.The introduction of ITQs results of a least squares approach is will break the continuity of the time given. Time did not allow bootstrapping seriesby changing the behaviour of forerror analysis. the fishers. It was strongly recommended that fishery Analyses to estimate ecological independent surveys of relative sustainability indicators have been abundance, be conducted both pre attempted priorto theworkshop, but the and post-ITQ implementation. This findings of this work was confirmed in was seen as the highest priority for the workshop. Lack of understanding of this fishery. the complex dynamics of the resource means that interpretation of catch rate, c. Since growth rates arestill the size-frequency,sex ratio, tagging, source of greatest uncertainty in our dredge and other sources of data are knowledge of spanner crab biology, difficultto interpret. further study needs to be undertaken to improve the understanding of the 2.4.2. OVERALL PROGRESS TO DATE growth of spanner crabs. Normal A summary of progress to date in terms fishinggear only captures larger of data quality, research and stock animals and thus tagging assessment knowledge is given in experiments would not include Table 2.4. (see over) juveniles. Without information relating to the initial elements of the 2.4.3. MONITORING, RESEARCH growth curve (involving juvenile DIRECTION AND PRIORITIES crabs), it is only possible to estimate a. Currently, spanner crabs are L°" but there is insufficient monitored solely through the use of informationto estimate the growth commercial catch rate data. During rate parameter K with any precision. the workshop, there were many Although the dredge gear only doubts expressed concerning the caught low numbers of juvenile validity of using simple catch rate animals, the data improved the data as an index of stock biomass. growth curve estimation. Given the

27 improved growth curve, approximate d. The possibility of using lipofuscin estimates of natural mortality could (a brain pigment which accumulates be made. Continued dredging for with age) as an ageing method juvenile crabs was considered to be should also be investigated as an essential by the Workshop. alternative approach to generating a precise growth curve,

Category Comments. Commercial Catch Relatively reliable. Effort Relatively reliable. Effort as dilly lifts less reliable prior to 1994 Catch rate Standardisedcatch rate by region. Assumption that catch rate proportional to biomass under debate. Recreational Catch, effort Probably not significantin comparison and catch rate with commercial catch, but has not been not estimated. Independentindex of biomass Commenced 1999. Estimates of natural mortality Only highly preliminary analysis completed in workshop, but not accepted due to ad hoe nature of estimation procedure. Difficult to measure, as the animal has not been aged, length frequenciesare highly variable and growth rate is still unknown. Estimates of fishing mortality or None. biomass Input controls Limited entry, size limits, spawning closure. Output controls Two Management Areas. Management Area "A" has TACC, daily bag limit and non-fishingdays. TACC Decision rules Objectively tested and performs relatively robustly under adverse conditions. Performanceindi cators Included in above decision rules.

Table 2.4. A summary of progress in research, data and stock assessment modelling.

28 2.5. REFERENCES Brown, I.W. Kirkwood, J., Gaddes, S., Dichmont, C.M. & Ovenden, J. 1999. Population dynamics and management of spanner crabs Ranina ranina in southern Queensland. FRDCReport 95/002. Chen, Y. and Kennelly, S.J. 1999. Growth of spanner crabs, Ranina ranina, offthe east coast of Australia. Journal of Marine and Freshwater Research. 50(4): 319-25.

29 30 common name EASTERN KING PRAWN

Penaeus plebejus

31 3.1. PARTICIPANTS

Dr Ian Brown (Fisheries Scientist, DPI, Deception Bay) Mr Rik Buckworth (Fisheries Scientist, NT Fisheries, Darwin) Ivlr Michael Cosgrove (Fisheries Scientist, DPI, Deception Bay) Dr Tony Courtney (Fisheries Scientist, DPI, Deception Bay) Ms Cathy Dichmont (Fisheries Scientist, DPI, Deception Bay) Dr David Die (Fisheries Scientist, CSJRO, Cleveland) Mr Mike Dredge (Industry Manager, DPI, Deception Bay) Mr Peter Gaddes (Trawl fisher, Brisbane) Dr Neil Gribble (Fisheries Scientist, DPI, Cairns) Dr Malcolm Haddon (Head, Department of Fisheries, AMC, Launceston) Mr Simon Hoyle (Fisheries Scientist, DPI, Deception Bay) Ms Kath Kelly (Fisheries Scientist, DPI. Deception Bay) Dr Steve Kennelly (Fisheries Scientist, NSW Fisheries, Sydney) Dr David Mayer (Biometrician, ARI, DPI, Yeerongpilly) Mr Michael O'Neill (Fisheries Scientist, DPI, Deception Bay) Dr Ian Poiner (Fisheries Scientist andTrawl SRC Representative, CSJRO, Cleveland) Mr Peter Seib (Trawl fisher and Trawl SRC/MAC representative, Sunshine Coast) Dr Ian Smith (Trawl Fisheries Manager and Trawl SRC Representative) Mr Dave Sterling (Trawl fisher, Brisbane) Mr Clive Turnbull (Fisheries Scientist, DPI, Cairns) Dr Reg Watson (FisheriesScientist, WA Fisheries, Perth) Ms Kate Yeomans (Fisheries Scientist, DPI, Deception Bay)

32 3.2. INTRODUCTION weeks before undertakingan extended seaward,northerly migration. 3.2.1 BIOLOGY An early tagging study (Ruello 1975a) 3.2.1.a. General Biology and suggested that there was a single adult Distribution population consisting of prawns from The easternking prawn, Penaeus many estuarine habitats. This was plebejus, is endemic to the east coast independently supported by allozyme of Australia fromcentral Queensland studies that showed genetic homogeneity (20°S) to north easternTasmania (42°S) for samples fromsouth-east Queensland ° ° (Kirkegaard andWalker 1970; Ruello (27 S) to Victoria (38 S) (Mulley and 1975a). It is the largest of Australia's Latter 1981). In NSW, additional endemic penaeid prawns in the genus tagging experiments confirmedthe Penaeus; femalescan reach 300 mm northward migration and mixing of total length and exceed 150 g (Grey et prawns, thus supporting Ruello'ssingle al. 1983). Juveniles occur in shallow stock concept (Montgomery 1981, 1990). embayments and estuaries, while adults It has been argued that, for stock commonly occur in oceanic conditions assessment purposes, two substocks to depths of 250 to 300 m. Commercial exist based on the origin of recruits. fishing is restricted to the continental These were referredto as the Moreton shelf and shelf breakfrom Lakes Bay-Mooloolaba substock, which had Entrance, in Victoria, north to the Swain recruits principally fromMoreton Bay, Reefs adjacent to the central Queensland and the NSW-Southport-Mooloolaba coast. In the central Queensland region substock which derived recruits the shelf and thefishery extend approx principally fromNSW estuaries. imately 200 km out fromthe coast. Elsewhere, the shelf narrows and the The two substock theory was challenged spatial distribution of the fisheryis by the expansion of the fisheryin the limited to a maximum distance of mid 1980s to include fishing grounds approximately 60 km fromthe coast. further north and offshore,near the Swain Reefs (22°S). Preliminary Penaeus plebejus display a seaward and examination of mitochondrial DNA northward migration of sub-adults as fromP. plebejus and otherpenaeid they mature. Adult females releaseeggs prawns fromAustralian coastal waters, that sink to the bottom and hatch within including the SwainReefs, indicate that about 24 hours, producing a pelagic P. plebejus has low genetic variation larval stage. Larvae remain in the water compared with the other species, with column for 3 to 4 weeks, using tidal and no clear spatial patternof genetic diel behaviour to assist them undertake a differentiation (Lavery & Keenan 1995). general shoreward migration. Postlarvae It has been suggested that the results then enter the benthic phase of their life were consistent with the highly cycle, settlingon shallowcoastal nursery migratory behaviour of P. plebejus. grounds, where they remain as juveniles for1 to 2 months beforecommencing a Although the easternking prawn fishery seaward migration to grow, mature and is generally considered to be a single reproduce. Penaeus plebejus differs stock, in the past it has been compart frommost other prawns in that it is more mentalised for modelling and assessment oceanic and migratory, remaining in purposes in alignment with state shallow estuarine areas foronly a few boundaries.

33 1 3.2.1.b. Growth and Mortality week for 1979 and 1980 respectively Several scientists have described the have been published, but these were growth rates of easternking prawns. biased upwards as they include other Lucas(1974) and Glaister et al. (1987) mortality sources (Glaister et. al. 1990). fitted von Bertalanffygrowth curves to 3.2.1.c. Migration and Movement data obtained fromtag-release experiments(Table 3.1). Lucas' In southeast Queensland, the emigration estimates of the growth coefficient (K) rate forP. plebejus migrating from Moreton Bay to adjacent offshore waters for males were higher than those 1 (E ) produced by Glaister et al (1987), was high = 0.17 week- (Lucas 1974) - about 4 times the fishing mortality rate possibly because Lucas tagged smaller, 1 faster-growing individuals. The quantity (F = 0.04 week ). When the combined to is generally assumed to be zero as effects of emigration and mortality were tagging data does not give an indication considered, an initial population in the of to. Growth is modelled from the size Bay was reduced to about half in two at recruitment, estimated at weeks. Several studies suggest that approximately 20 mm CL. Ruello P. plebejus utilise nursery areas and (1975a) also used tagging data from estuarine embayments for only a few

1 Reference K (week- ) L..

MALES FEMALES MALES FEMALES Lucas (1974) 0.10 0.10 40.0 mm CL 49 mm CL Glaister et al. (1987) 0.0595 0.0483 45.4 mm CL 59.5 mm CL

Table 3.1. Von Bertalanffy growth parameter estimatesfor P. plebejus.

experiments conducted on the NSW weeks before undertakingmigrations coast. He found growth rates were to deeper, oceanic waters. similar to those of Lucas(1974) but could not produce a growth parameters A compartmental model has been due to insufficientdata. Somers (1975) developed of the northward migration used both monthly length frequency of eastern king prawns in NSW. They distributions of post-larvae and juveniles concluded that emigration rates varied (2.5 mm CL-11.0 mm CL), and tag with location, and may actually reduce release data from prawns larger than with increasing water temperature 19 mm CL. He concluded that growth (Gordon et. al. 1995). of post-larvae and juveniles could be Catchability of easternking prawns described exponentially and that the varieswith lunar phase, however, sub von Bertalanffy expression adequately adults and adults appear to be affected described growth in the larger prawns. differently. Inestuaries and shallow An estimate of the instantaneousrate oceanic areas (< 30 m) catch rates of sub of natural mortality M forP. plebejus adults increase in the period leading up to in Moreton Bay and offshore waters andincluding the new moon, while inthe adjacent to Mooloolaba is 0.11 week-1 deeper oceanic areas(> 100 m) catch 1 and 0.05 week- offshore(Lucas 1974). rates of adults peak shortly before the full Other estimates, of 0.0621and 0.0810 moon and decline over the following

34 seven days. Consequently, the activity (Young and Carpenter1977, Young patternsof traw 1 operators targeting 1978). Although Young and Carpenter easternking prawns in shallow ( < 30 m) (1977) concluded abundance peaked and deep (> 100 m) water regions of the between July and September in Moreton fishery differaccording to lunar phase. Bay, post-larvae were abundant year Fishers operating relatively small vessels round and seasonal trends were weak. in shallow water preferfishing in the The aversion that P. plebejus exhibits to period leading up to andincluding the areas with a freshwater influencewas new moon, while those operating larger supported by Coles and Greenwood vessels targeting adults in deeper waters (1983) who found that, in the Noosa preferfishing in the period leading up River system (approximately 150 km to andincluding the fullmoon. The north of Moreton Bay), post-larvae only catchability of adult males andfemales settled at sites near the river mouth, and in offshorewaters (160 m) also changes only for brief periods. over the lunarcycle. As a consequence, therelativ e contributionto the catch from 3.2.1.d. Fecundityand Spawning each gender differsover the lunar cycle. The spawning stock dynamics of P. Lunar phases also affect the incidence of plebejus are poorly understood. Dakin mature and ripe females. During the (1938) andRacek (1959) used field main spawning season, two periods of observations of the distributionof increased spawning activity occur in inseminated adult females, eggs, and each lunar cycle. Lunar phaseeffects larval stages to infer reproductive activity. need to be considered whenever indices Racek (1959) observed the population of spawning stock size are examined between 27°S and 36°S and suggested the (Courtney et. al. 1996). 'period of maturity' was fromMarch to June and thatbreeding grounds were in Adults are oceanic andare among the depths of 100 to 140 metres, butwarned most migratory of the Crustacea. his results were inconclusive due to Several tag-release studies in NSW and difficultiesin identifying larvae tospecies Queensland have confirmeda general level. Laboratory experiments indicated northerlymigration along the coast and spawning and maximum hatching success into deeper water (e.g. Ruelle 1975b; for P. plebejus are likely to occur in Glaister et al. 1987; Potter 1975). oceanicsalinities (30-34ppt). Based on Tagged individuals have been recaptured the recapture of tagged prawns, Ruello over 1000 km fromtheir release (1975a) suggested the coastal area location. This has been referredto as a between Fraser Island andSouthport spawning migration, emphasising the (Figure 2.1) was the most important importance of the northernparts of the spawning area forthe species. However, fisheryas the source of spawning stock, thiswas prior to the establishment of however, recently the significance of additionaltrawling grounds for this such a spawning migration have been species north of about 26°S. questioned. Fecundity ranges from350 000 eggs for Barber and Lee (1975) and Rothlisberg 40mm CL females to 885 000 eggs in 60 et. al. (1995) have shown planktonic mm CL females. When the effects of larval stages of P. plebejus enter growth, mortality and fecundity are Moreton Bay with the flood tide during considered simultaneously, the size both day and night. Post-larvae typically range of femalesthat are responsible for settle on bare substrates and seagrass most egg production is 35-48 mm CL. areas, with fewerindividuals settling in Large females (>50 mm CL) as a whole areas that have a freshwaterinfluence

35 contribute little to egg production November. Individuals recruit at the because there are very fewsuch large relatively small size of 14-15 mm CL. individuals left in the population, due to Selectivity ogives fora 1 5/8" mesh the high mortality rates. The capacity of net indicate that very fewindividuals very large individual females(� 60 mm (<10%) are retained in the gear. CL) to produce and fertiliseeggs is Recruitment to the fishery occurs reduced apparently by reproductive 1-2 months afterthis period as senescence (Courtney et. al. 1995a). individuals grow and migrate offshore. Approximately 50% of 19 mm CL The spatial distribution of mature individuals are retained. females was investigated in southeast Queensland using measures of ovary The main abiotic factor affectingthe weight and histological condition. distribution of recruits in Moreton Bay is Ovary weight and the incidence of depth; recruits are negatively correlated mature or ripe females did not vary with depth. Two areas in Queensland are significantlybetween areas. No obvious considered to be majorsources of temporal patternsin spawning activity recruitment; Moreton Bay and the Great were apparent. The main spawning Sandy Strait-Wide Bay Bar region. The period appears to be June (winter). relative contribution to recruitmentin Mean ovary weight measures, and the Queensland fromNSW is unknown incidence of histologically mature (Courtney et. al. 1995b). femalesare slightly elevated at this time. 3.2.2 However, the main spawning period of DESCRIPTIONOF FISHERY June has been largely deduced by back 3.2.2.a. Introduction calculating fromthe time of thepulse In the order of 200 vessels land 1,500- recruitment using the growth curve and 2,000 tonnes of easternking prawns age at length estimates (October annually in Queensland. The catch is November each year). Thereis likely to restricted to the south-east comer of the be a significantlevel of egg production state. An additional 800-1 000 tonnes outside the main spawning period that arelanded in NSW annually. Retail does not contribute to stock renewal prices for the product vary with prawn (Courtney, 1997). size and season. Small or 'Bay' eastern The significance of the northerly spawn king prawns (100 per kg) retail forabout ing migration has been challenged. $6-8 per kg, while larger 'Ocean kings' Given the low speed of cross-continental (10 to 20 per kg) retail for about $20 per shelf currents, it seems that many of the kg, although ocean king prawns retailed larvae spawned offshore and in the for $40 per kg over the 1997 Christmas northern regions of the fishery have little season in Sydney markets. The majority chance of renewing southern coastal of the product is sold on domestic populations. It has been suggested that markets, mainly in Sydney. A very localised near-shore spawning stocks minor component of the overall catch is NSW. may be more important to the stock landed in a recreational fishery in renewal process than firstconsidered InMoreton Bay the annual pulse of (Rothlisberg et. al. 1995). recruits is fished heavily from November 3.2.1.e. Biological Recruitment to January. P. plebejus migrate rapidly through the Bay to deeper oceanic areas. The annual timing of biological The range of mesh sizes allowed in the recruitment in Moreton Bay is distinct Bay is 38-60 mm, with most vessels and restricted largely to October- using the minimum size allowable.

36 Maximum head rope length is 32.5 m, The Moreton Bay fleet tended to fishin (deployed usuallyas twin gear) and the bay in summer andthe deep-water maximum vessel size is 14 m. Depths fisheryin winter. The fleetrapidly grew range from4-35 m with trawl shot in numbers as the northerntiger and duration of 1-2 hours. banana prawn fishery developed. The inshore nature of the fishery with calmer Outside the Bay, larger mesh size (50 waters and protected anchorages lent mm, 2") is deployed. Maximum vessel itself to fast development. At that time, size in the offshore fisheryis 20 m. governmentwas giving encouragement Trawling occurs in depths to about 300 by way of a 25% investment allowance m and shot duration increases to 3-4 on the construction of new boats. The hours. In the Swain Reefs region, red spot king and bug fishery had not fishing trip duration increase to been established at that time. approximately 12 days beforereturning to port. Logbook data indicate average As the fleet grew, so did the average nightly catches in theorder of 50-120 kg horsepower and boat size, net size and per vessel with strong seasonality. catching efficiency. The catching ability of a modernprawn trawler relative to its From December to May the annual earlier counterpart, is far greater than we pulse of recruits is fished down; monthly realise. Sonar, GPS and plotters make it effortand catch decline over this period. possible to catch prawns in areas that The spatial distribution of effort moves were once protected by reef and offshore and northward, following the obstacles. Once a concentration of migration of prawns. By May, natural prawns is foundit is easily pinned down and fishingmortality rates dictate that to the exact co-ordinates with the aid of only about 5% of the annualrecruitment a try net and GPS with plotter. Mr remain to contribute to spawning in June. Gaddes, a professionalfisher, believes thatthe increase in catching efficiency is 3.2.2.b. The Development of the Fishery highlighted by the factthat the catch rate The development of the fisherywas of a concentration of prawns now well described by Peter Gaddes. His declines rapidly over one, possibly two involvement with the prawn fishery nights working. Increased horsepower startedas a deck hand in 1972 in the and the use of multiple net rigs (e.g. 4 Moreton Bay and adjacent deep-water nets and try net or 3 nets and try net) fishery offCape Moreton. At that time, have also contributed to the increased a 14-metre trawler was considered to be catching efficiency of the fleet. The a large boat. OnMoreton Bay, one 8 duration of a trip has been greatly fathom net or two 4 fathom nets were increased by the introduction of freezer allowed as is still thecase today. In the rooms and largerhull capacity, most offshorefishery, one net between the boats being able to achieve two weeks sizes of 10 to 15 fathom was used at sea and some even 4 weeks. depending on the horsepower of the The unfortunate reality in this fishery vessel. There were fewrefrigerated is that increased catching ability has boats and ice was mostly used keeping not meant increased profit, because the the maximum time spent at sea to below stocks are depleted and running costs four days and even shorter if the wind are increasing. This declining trendis reached 20 to 25 knots. At that time, he reflected in the fact that more new did not recall any boats withradar. An trawlers with largefuel capacities are unsophisticated echo sounder was the being built. More fuel means moretime only hi-tech piece of equipment on board. at sea.

37 The geographical distribution of the Nearbar closures have been used as a eastern king prawn fisheryis from form of spatial yield optimisation in Sandy Cape to the Queensland-NSW Southport, Jumpinpin, Amity and Tincan border. In November, new recruits begin Bay. At present, there are no clear to be caught in the shallows (40 tolOO targets, reference points or measures metres), close to the bars and northern directed towards achieving the Act's outlet of Moreton Bay e.g. Cape Moreton objectives but these are being developed to Mooloolaba. These prawns migrate in a new Management Plan. Key issues into deeper water as they grow and are that are being addressed in this fished from March through to December Management Plan include: in the 100 to 200 metres depth range. • The current vessel replacement Scallops are fished all year roundin policy discourages vessel Hervey Bay and offYeppoon. The trend replacement and has resulted in the o to date is f r boats to migrate south into average vessel being 25 years old; the scallop and king prawn fishery while the tiger grounds are closed. • There is an enormousamount of latent effort. The fleet of about800 3.2.2.c. Management boats now apply about 90 000 The objectives of the Fisheries Act night's work, which could be readily 1994 (Qld) areto ensure that fisheries increased by 30-40%; resources are used in an ecologically • There is a belief that the industry is sustainable way, to achieve optimum runningunder a very low profitregime; community, economic and other benefits • obtainable fromfisheries resources, andto The stocks are fully exploited and ensure access to fisheriesresources is fair. chronic recruitment overfishing is a possibility. The easternking prawn fishery forms part of an overall east coast otter trawl The key elements of the proposed Draft fishery, which is managed by restricted Management Plan are: entry (about 800 boats). Other than size • Restricted access, but under certain restrictions applied to boats in Moreton conditions, all licensed trawlers can Bay, all trawlers are entitled to access all access all traw 1 resources; trawl species. A further 200 odd licence • A complex effort capping endorsements allow small boats ( < 10m) mechanism which involves boat to use beam trawl in specific estuarine units and historicalperformance; areas of the state's waters. Eastern king prawns are fished sequentially, fromthe • Seasonal closures in north and south inshore nursery areas to the offshore Queensland; adult animal. There are different gear • B ycatch reduction devices and a restrictions forthe inshore and the turtle protection mechanism; offshore fleet. Beam trawlers in the rivers are restricted by a maximum 5 • Limit reference points; if catch rates metre beam and footrope and the close decline below 70% of the long-term inshore fleet, in the bays, by a maximum average, within key life-cycle of 8 fathoms of headrope. The inshore periods, then spatial closures are fishery (less than 100 metre deep) and triggered. Ifcatch rates decline below offshore (greater than 100 metres) are 60% of the long-termaverage a more limited to a maximum of 88 and84 significant response is proposed. metres of headrope andfootrope Compulsory VMS for monitoring will be respectively. phased in, starting with the scallop fleet.

38 3.3. EASTERN KING Common Name Elsewhere Moreton Bay PRAWN TASK GROUPS Banana 18.36 3.52 Bay 1.71 48.53 1. Characterisation of catch rates in Clicker 0.00 0.14 the Fishery: Coral 3.57 0.22 Eastern king • amalgamateNSW and 0.00 0.00 Endeavour 3.21 1.38 Queensland data; Greasy 0.01 1.71 • investigate the influence of Hardback 0.00 0.05 geographical and temporal scale King 62.80 25.28 on catch rates; Leader 0.01 0.00 • regionalisation Mixed 0.39 5.35 Mixed bait 0.09 0.40 2. Distributionof effort: Pink tailed 0.01 0.00 • development of thefishery; Pistol 0.00 0.01 • level of reportingscale in termsof Red spot king 0.69 0.01 6 minute and 30 minute grids. Royal red 0.09 0.00 3. Evidence and arguments for Sand 0.01 0.00 Scarlet recruitment or growthoverfishing. 0.08 0.00 School 0.56 0.50 4. Establishment of a recruitment index Tiger 8.03 12.28 through simulation models. Unspecified 0.37 0.62 Western king 5. Location and size. 0.00 0.00

3.3.1 CHARACTERISATION OF CATCH RATES Table 3.2. The relative proportion of the NSW catch and effortdata was very differentprawn species/categories in the o CFISH data-base inside Moreton Bay and kindly made available f r this workshop. Elsewhere (south of 20°S). The zero values are For confidentialityreasons, this data either truly zero or an extremelysmall amount was summarisedby month, unlike the landed. Note that the species category Eastern Queensland logbook data, which was King Prawn receives no use. at the record level. This made it impossible to combine the data without Coast Otter Trawl Fishery Logbook. losing the resolution necessary foreff ort This is used to represent eastern, red characterisation. In this group, there spot and western kingprawn (Penaeus fore, work was restricted to the latisulcatus). The problem of 'bay Queensland data-set. prawns' was considered to be mostly There was much discussion about data confinedto Moreton Bay (Table 3.2). qualityand the difficultiesin separating The difference between Moreton Bay the species of interest fromone another. landings and landings fromelsewhere A major source of noise and erroris the (but south of 20°S) are primarily related category 'bay prawns' which can contain to differencesin bananaprawns, 'bay' any smallprawns including easternking prawns and 'king' prawns (Table 3.2). prawns, greasy back prawns (Metapen It was clear that Moreton Bay and aeus bennettae), school prawns (M. elsewhere should be treated separately macleayi) andred-spot king prawn (Penaeus longistylus). Fishers often to otheroceanic waters. classify prawns as 'bay prawns' when Owing to the problems in identifying the they have not sorted their small prawn easternking prawns within the database catch. A further problem stems from the in the time available it was decided to category 'king' as used on the East

39 limit analyses to relatively simple the offshore vessels are larger and have summary methods. Given the doubts higher catch rates. about the nature of the data they were working with, the group recommended InsideMoreton Bay, the twelve-month that the data-set be cleaned using an moving average withinyears was less agreed upon algorithm designedto select than expected. Across the period1988 to records which would be mostly eastern 1997 therewere relatively low catch king prawns. For example, it was rates in 1993 (Figure 3.1). On a shorter suggested that the category 'king' could temporal scale, there were clear peaks in be divided into easternking and red-spot catch rate over new year periods king prawns using locations defined by (November, December and January latitude. To deal with the problem of which were consistent with the expected 'bay prawns', especially in Moreton periods of recuitment. However, there Bay, the group recommended that such was also an unexpected peak in catch landings be sampled to determine the rates occurring in mid-year (June and proportional species composition. July) which was of variable magnitude (Figure 3 .1 ). This mid-yearmaximum As the data did not lend itself to detailed had not been mentioned before and the standardisation, it was decided to group feltit was an interesting and characteriseseasonal changes in catch repeatable phenomenon that needed an rates on a broad spatial scale (inside explanation. (The FRDC 97/145 Project Moreton Bay and elsewhere). At the 'Developing indicators of recruitment same time long term trends were and effectivespawner stock levels in considered by using twelve month easternking prawn' will investigate this moving averages to remove the strong phenomenon further). The suggestion seasonal signal fromthe data fromboth that it was a reflectionof spawning inside Moreton Bay and elsewhere. aggregations was countered by the This group did not have details of the statement that the observed peaks were characteristics of the vessels in the too small. Another suggestion was the easternking prawn fishery, however, mid-year peak could be simply a

40 - 35 ''• �. 30 ' +:'' ,, '' ::e ', 25 Cl) 20 -; , . 15 • -; 10 •• (.) 5 0 Jan- Jan- Jan- Jan- Jan- Jan- Jan- Jan· Jan- Jan- Jan- 88 89 90 91 92 93 94 95 96 97 98 Year Figure 3.1. Catch rates of the categoryking prawns inside Moreton Bay (points with dotted line) with a twelve month moving average (solid line). Note that there is a peak every January/Februaryand another everyJune/July.

40 reflection of that time having thebest analyses above identifieda number of weather conditions for trawling. It was interesting patternsof activity within the also pointed out that the 30 minute grid fishery. The group recommended that block used to define Moreton Bay following data clean-up, as described includes a small offshoresegment and above, the mid-year spike in catch this might influencethe results observed. rates within Moreton Bay should be investigated further. The easternking prawn catch rate outside Moreton Bay is also clearly 3.3.2 DISTRIBUTION OF EFFORT affected by season (Figure 3.2). Catch Theobjective of this group was to rates peak over the new year investigate thespatial resolution of the corresponding with recruitment and logbook data and how the spatial subsequent decline through to late distribution of effort has changed over winter. It was suggested that there may time. There was a known expansion in be additional recruitment to the fishery thedistribution of the fishery during the fromsouthern immigrants moving north. middle 1980s. However, this occurred Outside Moreton Bay there was no beforethe logbook database was evidence of a mid-year spike in catch established. Theresolution in the data-set rates as observed inside the bay (Figure is low, mainlyreporting at 30 by 30- 3.2). The twelve-monthmoving average minute grids, but has improved over time showed a clear and smooth trend that to 6 by 6-minute grids. Some precise was flat through to 1994 with a peak in latitudinal and longitudinaldata are being 1995 and 1996 which then declines to logged. An animated EXCEL sheet that the 1988-1994 level. scrolled through the effortdistribution by depthand latitude was created. An Even though the catch rate data for example of 1989 and 1997 effortis given easternking prawns have many problems, in Figures 3.3 and 3.4. the information obtained from the simple

140 ------,---,---, _ 120 :e 100 - 80 60 40 20 4----i----+--+--+----+----+--+---r-----t----' 0 Jan- Jan- Jan- Jan- Jan- Jan- Jan- Jan- Jan- Jan- Jan- 88 89 90 91 92 93 94 95 96 97 98 Year Figure3.2. Monthly commercial catch rate of categoryking prawns caught outside Moreton Bay. The solid line is the 12-month moving average.

41 1989 ,,---,---,.-..--,--r-.---,---,---,,---,---,---,,---,--,--,--,---,--i20 111600-750 =4---1--1---+--l---l-+--+----+-+--+---t-+--+---t- 111450-600 21 m!300-450 -- - - + - - 22 'iii' �-1--+-�1 +---1---t +--t---+ t---+-----ttt--+--+ t- III 150-300 a, '> ! ,,.-+---1---t-+--+--+-+--+--+-t---+---i-- 0 0-150 <.. 3 0) a, =t---+--t---11--+--t---l-+--t---t-;--;--,--t---i----i 248. ,,._t---f--t---11---f--t---l---t--t----1�;--;--,---r---i 25 "g �-+--t---+---1 26 5 R,Hl--+--t---+---12 7 =='-'---'---'---L--'---'---'--..__.....__.__..__, 28 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Depth (*10 m) Figure 3.3. Distribution of effort, as number of boat nights, by latitude and depth/or eastern king prawns in 1989, excluding Moreton Bay.

1997 r--"tn"--r--ir-..,..,.,-,,,_�,r,---,--y---,--,---.---.--,---.--r----�20 11600-750 �-l--4-��+--+---1--+--l--+-+---+--+-+---+--+-1-- l!ffl 450-600 21 � I---- l:ill 300-450 -.. ---' + 2 2 D 150-300 --- l t--+---+- �-+--+-t--+--+-t---t--+---,t----t--+ t--,-- D0-150 23 ..,, a, �+--+---¼-+--l--+-+---+--+-+---+--+-+---+--+-r--1 24e, a, 1,/--l--t-t--+--+-t--+--+-t--+--+-t--'1---t---1 25 � �:'tl--+--+-+--l 26 ii ...I ==+---1--+--l---1----12 7

l,__...L.,;;...c...i..;...;__L-....1....--l---1--'---'---'---'---'---'---'--_.___,__..__.....___.__..__, 28 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Depth (*10 m)

Figure 3.4. Distribution of effortas number of boat nights, distribution by latitude (degrees) and depth (m)for easternking prawns in 1997, excluding Moreton Bay.

In the two figuresabove, about 3 3 % of Using the distribution of catch by the data include zero andnull depth or month, the group identifiedthree areas latitude values. The Moreton Bay effort (Figure 3.5): data are not included. The extremely high a. shallow recruitment (Moreton Bay); effortwithin thisregion would remove b. offshorespawning (>100mdepth); contrast in the non-Moreton Bay areas. c. intermediate areas ( <100m depth Withthose records having a null depth excluding Moreton Bay). and precise latitude and longitude, it should be possible to inferdepth. There is a clear reversal of seasonal Unfortunately, the group did not have the catch levels between offshorespawning time to attempt to include such records in areas and elsewhere. Catches peak in their analyses. The group suggested that Moreton Bay and the intermediate areas this work be pursued when possible. during the new year period, there is

42 1400 O > 100m 1200 � <100 excluding Moreton Bay - 1000 -+- Moreton Bay C -C 800 -r:. 600 400

200

0 1 2 3 4 5 6 7 8 9 10 11 12 Months

Figure 3.5. Catch levels of easternking prawn in different areas and depths. The changing seasonalitywith location, relating to migrationwith growth, is clear.

some evidence of catches being 3.3.3 EVIDENCE OR ARGUMENTS maintained for a longer time in the FOR RECRUITMENT AND/OR intermediate areas.The offshore GROWTH OVERFISHING spawning adults exhibit the highest catch The group used thequalitative ranking levels in the mid-year winter months process by Penn (1984), which considers (Figure 3.5). This patternreflects the the life history characteristics of each offshoremigration of small prawns as species to determine whether it will be they grow. This could be confirmed by vulnerable to overfishing; the examining these patterns through a implication was that P. plebejus has a season. It was considered by the group low riskof overfishing. This is due to that such seasonally structured maps its highly dispersed nature, that it is only would also illustrate the general active (and fished) at night and that it northwards movement of small prawns seems to be buried through most of the as they grow. night, a behaviour most likely to reduce The group made the following its catchability. recommendations: Logbook data from two possible sites a. When mapping effort distribution, were suggested for investigation of effort should be sub-divided by information on a recruitment index. The season; first is Moreton Bay, but the inability to b. Estimate where possible, null depth distinguish between bay prawns and valuesby using latitude and easternking prawns in the database longitude data; makes analysis difficult. The second area is the Wide Bay Bar area (Grid c. The offshore spread of effortshould W34). The average of the catch rates for be linked to migrationand tagging January and February were used as a information, using both NSW and recruitment index, consistent with life Queensland data. history patterns.This index does show a

43 decline over time. Ifsome indirect in 1990 resulted in an increase of 10%. attempts at effortstandardisation are On top of these increments , a yearly attempted, then the decline becomes geometric effortincrease of 1 % was more marked (Figure 3.6). After the also assumed. workshop, it was suggested that recruit '5%' : Same as for '1%' but higher ment might occur over the whole of increases. Thirty percent for Quad gear, October through to the end of January 10% for try gear and 15% increase for and that the average across these dates GPS. Also, the overall geometric would provide a better recruitment increase on top of the incremental index. changes was assumed to be 5% per year. The three series of catch rates were The 5% series was clearly the worst case produced with the following assumptions: scenario. The industry members in the 'Norn' : Unstandardised mean January workshop generally agreed to the and February catch rates. figures. The consensus was that the true 'l %' : Triplequad gear was introduced scenario would be between the 1 % series in 1976 to 1978 and introduced an and the 5% series. Factors not included estimated 20% increase in efficiency to would be recruitment changes associated the effort. In the mid-1980s, the with climatic changes, random noise and introduction of try gear was thought to changes in broodstock availability, for increase effectiveeffort by 5% and GPS example, due to recruitment overfishing.

w 200�------,Quad gear GPS -••-Nominal ::::, Try gear · · • · · 1% p.a. a. 0 160 B 5% p.a.

G) 120 0 0 80 ...... ::::, D. 0 40 0 1977 1982 1987 1992 1997 Year 1 Figure 3.6. Easternking prawn corrected and uncorrected catch rates (kg. boat nights- ) of Januaryand February from1977 to 1981 (voluntary logbook data collected byM. Dredge, DP/, SFC, Deception Bay) and from1988 to 1997 (CFISH database, QFMA)for Tincan Bay. Arrows indicate years in which a technological improvement, assumed to effect fishing efficiency, was addedto the fleet. An optimistic and pessimistic cumulative annual effort increase of 1% and 5% has also been added.

44 3.3.4 ESTABLISHMENT OF And January A RECRUITMENT lNDEXTHROUGH SIMULATIONMODELS A monthly least squares biomass Due to differences in the catch rate dynamic model was fittedto monthly datasets between NSW, Queensland and catch rate data fromQueensland alone, both states combined, theequations NSW alone and a combination of the above have to be adjusted to align with two data-sets. The available data set and the observed recruitment periods for recruitment patternsof the prawn differ each State or combinations of States. by state, so the NSW model and the Queensland model start in October 1985 The relationship between catch rate 1 and 1989 respectively. fu both cases a (kg. boat nighf ) and biomass was strong seasonal trend is apparent. In modelled as: Queensland, the highest catch rates (and, it is assumed, peak recruitment) occur fromDecember to March. The pattern is - =Bq y,m : 3.3 slightly different in NSW, with highest (CJE y ,m catch rates being observed between January and March. where Ey,m is the effort in year y, month m, and A simple variation of structure permitted q is the estimated catchability the behaviour of two models to be coefficient. investigated. The first model assumes that the increase in catch rates is due to A recruitment index was estimated for recruitment and that the decline in catch each year. A single global catchability rates is due to commercial catches only. parameter was estimated assuming catch The second model has an additional term rates were directly proportional to to reflectthe changes in biomass due to biomass. A further parameter, initial natural mortality and individual growth. biomass, needed to be estimated. In the Recruitment is uniformly distributed first model, the y parameterwas set to between the relevant months. one, cancelling its effect. In the second model, it was estimated directly. The The biomass dynamic equation for model fitto each data-set are given Queensland is: below (Figure 3.7).

February to December The resultant recruitment index forthe Queensland data shows a decline from 1990 to 1993 followedby a recovery until a maximum in 1995, followed once where Bt is the biomass (t) in year y, more by a decline. fuNSW, thereis an month m, initial period startingin 1986 when y is the growth and natural mortality recruitment is highly variable but from term, set at 1 in the firstmodel and 1990 recruitment is relatively stable estimated in the second, with only a slight peakin 1995. The Ct is the catch (t) in year y, month m, combined data was dominated by the • m is the proportion annual much higher catch rates in Queensland estimated recruitment that occurs in andexhibited an even more marked month m patternthan found in Queenslandalone Rtis the recruitment estimate in year y. (Figure 3.8).

45 120 -;;::: QLD "Cea 100 80 � - 60 w 40 ::, a. 20 0 0 co 0) 0 T"" N M <:I" LO (0 co (0 0) 0) cp 0)' 0) cp O') t5 t5 t5 t5 t5 t5 t5 t5 t5 0 0 0 0 0 0 0 0 0 Year

..-.. 90 ------.... �80 NSW+QLD "C 70 ._ 60 �50· l.� -40 \ : W 30 ::J 20 Q. 10 0------_, 0 co O') 0 T"" N M v LO co (0 (0 O') O') O') O') O') O') O') t5 ii t5 t5 is t5 t5 13 0 0 0 0u 0 0 0 0 0 Year

..-.. 90 �80 NSW "C 70 C) 60 � 50 -40 w 30 20 a. 10 0 0 'V LO co ,..._ (0 O') 0 N M 'V LO (.0 (0 (0 (0 co (0 O') a) 0) O') O') O') O') ..!, ..!, °i' t5 0 0 t5 t5 13 t5 t5 0 0 0 0u 0 0 0 0 0u 0 0u 0u 0u Year

Figure3.7. Model (solid line)fitto the Observed (dashedline) catch rates for the off shore fleetof i) Queensland (QLD), ii) Queensland and NSWfleet combined and iii) NSW

46 1.8 . � 1.6 . •,, ' >< \ • . \ • Cl) 1.4 • \ \ •. "C ' . \ \ C: ' \ - 1.2 ' '. ' ...., \ C: Cl) 1 •• E 0.8 ·-:::J ... 0.6 . . ·•· .. NSW a:Cl) 0.4 • QLD 0.2 ALL 0 86 88 90 92 94 96 Vear

Figure3.8. Recruitment indices for NSW, Queensland and both combined for the offshore fleets.

3.3.5 LOCATION AND SIZE fishingareas. These are the Swain Reefs DISTRIBUTION (U28, V28, W28, W27), Lady Elliott This group argued that if recruitment Island region (V30), off Mooloolaba overfishing occurs, then closures should (W36) and offshore of Moreton Island be designed to maximise egg production. (X37). This would be affected by abundance, Monthly size-frequency and depth data animal's size, timing of spawning and from both selected fishers and a DPI level of fishing effort. Pervious YPR research vessel were available for these studies on eastern king prawns have four regions. Size-frequency plots of indicated that the optimal egg production June and July prawn sizes at each size classes are females 35-45 mm of these regions are plotted below carapace length. This is due to a (Figure 3.10 a-d). combination of the fecundity-length Prawns in the Swain Reefs region are relationship andnumber of females alive extremely large and not many are within at a certain size (Courtney et al. 1995a). the optimal spawning size range class. Biological recruitment occurs in October As one progresses south, however, the to November. Based on the growth rates relative size decreases. Although of these animals, the effective spawning several animals between sizes 35 to period should therefore be between June 45 mm CL are present within the Lady and July. Ifthe commercial catch for the Elliot and Mooloolaba region, it is months of June and July are plotted in clear that the highest abundance of the GIS package ARCVIEW (Figure these females are outside and adjacent 3.9), it is possible to locate fourmain to Moreton Island.

47 EKP Catch (kgs)

0 - 5000 ... 5000 - 10000 10000 - 15000 15000 - 20000 20000 - 25000

June 1996 Catch

...

July 1996 Catch

Figure 3.9. ARCVIEW GIS graph of the distribution of June and July 1996 catch (kg).

48 25 >-- c.> VI 20 - Swains C .. Cl) Cl) ::::, .0 15 eer ::::, E 10 - u. ..s 5 0 - co C\l (0 0 v co C\l co 0 v co C\l co .,.... C\l C\l C') C') C') v v LO LO LO co co Carapace Length (mm)

40 >-- c.> VI Lady Elliot Island C .. 30 Cl) Cl) ::::, .0 er E 20 Cl) ::::, u.._ C - 10 0 co C\l (.0 0 v co C\l (.0 0 v co C\l (.0 .,.... C\I C\I C') C') C') v v LO LO LO (.0 (.0 Carapace Length (mm)

8 >-- c.> VI 6 C .. Mooloolaba Cl) Cl) ::::, .0 er E 4 Cl) ::::, u.._ C - 2 O· co C\l (.0 0 v co C\l (.0 0 v co C\I (.0 .,.... C\I C\I C') C') C') v v LO LO LO co (.0 Carapace Length (mm)

40 >-- c.> VI - East of Moreton Island C .. 30 Cl.I Cl) ::::, .0 er E 20 Cl.I ::::, u.._ -C 10 0 co C\l (.0 0 v co C\I (0 0 v co C\I (0 .,.... C\I C\I C') C') C') v v LO LO LO (.0 (.0 Carapace Length (mm)

Figure 3.10. a-d: Size frequenlydistribution of eastern king prawns fromJune and July survey data in 1991 and 1992 using 3 commercial vessels and 1 DPI vessel for a) Swain Reefs Area, b) Lady Elliot Island region, c) offMooloolaba coast and d) east of Moreton Island.

49 0.,------.------...... --. 3 • Deep • • -Q· • • Middle - 25 • Shallow Cl,) � 20 ::s C >- 15 C Cl,) ::S 10

18 23 28 33 38 43 48 53 58 63 68 Carapace Length (mm) Figure 3.11. Size:frequencyby depth of easternking prawns caughtoff Moreton Island during a sW1Jey in 1991 and 1992.

Ifthe size frequency distribution of Any potential closure should be easternking prawns in waters off carefully investigated in terms of risks Moreton Island is related to depth, then and benefits. A simple table was the animals between 35 to 45 mm CL produced by this group identifying these seem to occur in the medium depths of possible actions to be taken, their risk about 90 meters (Figure 3.11). In and benefits (Table 3.4). conclusion, any spawning closure should aim at the region offshoreof Moreton 3.4. CONCLUSION Island in depth greater than 85 meters during the months of June and July. 3.3.6 GROUP REPORTS Species identification problemsexist in In terms of growth overfishing,juvenile the database and algorithmsthat were closures would be necessary. The used in the workshop to isolate eastern obvious area would be Moreton Bay as king prawn (Penaeus plebejus) catch and this area is a major source of juvenile effort. The biggest problem exists in the easternking prawns. However, the Moreton Bay area,however 80-90% of social and political cost would be the fishery'scatch and effort occur enormousand, possibly, small areas or offshorefrom Moreton Bay. The catch night-time closures could be considered. rate data reflectsa distinctive pattern of Previous YPR work on these closures high values between November to has shown that assumptions about January, consistent with documented migration rates greatly affectthe periods of recruitment. This is followed duration of closures. Also, there is by a decline towards mid-year. A small generally only a 5-10% increase in value.

50 and unexpected mid-year peak of in 1995, followedonce more by a variable magnitude can also be decline. For reasons related to degrees observed. This may reflecta second of freedom, the present biomass relative minor recruitment pulse, possibly as a to virgin biomass was not estimated. It result of northward migrations from will be important to consider theeffect New SouthWales and is currently of effortcreep in the future development being investigated. of this model. The spatial resolution of thedata in the Commercial catches during the main commercial logbook system is coarse spawning period (June and July) were with most fishers reporting at the 30 by investigated. Based on prior per-recruit 30-minute grid scale. This has improved analysis, the optimal egg production size over time to 6 by 6-minute grids. The classes are females 35 to 45 mm larger grid level in Moreton Bay also carapace length. As a result, any includes a small offshore section. possible spawning closures to avoid Some attempt at using depth to further recruitment overfishing should aim at segregate the catch spatially was the region offshoreof Moreton Island in promising, but about 33% of the depthsgreater than 85 metres during the offshoredata included zero and null months of June and July. depth or latitude values. Three biological areas were identified, being shallow The working groups were unable to recruitment (e.g. Moreton Bay), offshore determine whether the easternking spawning (>100m depth) and inter prawn catches were or were not mediate ( <100m depth, but excluding sustainable. Attempts at effort ad hoe, thebay). standardisation were but demonstrated a significanteffect on A yearly recruitment index was catch rate. An objective method for established, using the average catch rates measuring effective effort changes for Januaryand February in the Wide should be utilised in future. Furthermore, Bay bar area. The historical data thevarying spatial and species resolution between 1977 and 1981 and the logbook of the data adversely affectsconfidence data from1988 to present were used. in the results. An historical database, Factors that may affect the catch rate currentlymanaged by QFMA, may be e.g. quad gear, try gear and GPS were extremelyuseful for extending the series considered and their possible effects back to 1977, but was not available to included in the analysis. An overall 1 % the workshop. Successful modelling and or 5% per year effectiveeffort creep management of eastern king prawns factor was added. Boththe worst case relies on good collaboration between andbest case scenariosof the index New South Wales and Queensland. demonstrated a decline in recruitment over time. 3.3.7 PROGRESS TO DATE A monthly least squares biomass A summary of data quality, research and dynamic modelled fitted to Queensland, stock assessment knowledge to date is New SouthWales and combined catch presented in Table 3.3. rate data was developed. The catch rate data were unstandardised. The model fitted the data well. The resultant recruitment index for the Queensland data shows a decline from1990 to 1993 followed by a recovery until a maximum

51 Category Comments Commercial Catch Problems in identifying eastern king prawn catch data within the Bay and King prawn CFISH categories. Effort As above. Catch rate CFISH data tends to have little year to year contrast. Within years there is a marked seasonal pattern. Historical data from voluntary logbook programs and other studies would extend the data to about 1970, but IS unavailable for analysis because the database entry is incomplete and unchecked. Data resides with QFMA. There is an urgent need to have the data developed and made available. Recreational Catch, effort Minor in New South Wales and and catch rate insignificant m Queensland. Not estimated. Independent index of biomass or Present FRDC project to establish recruitment appropriate methodology. Estimates of natural mortality Good estimates using various data sources in various studies, however few recent estimates. Estimates of fishing mortality or Good estimates using various data biomass sources in various studies. Workshop model estimated yearly recruitment. Input controls Limited entry, gear restrictions, inshore seasonal and area closures, hull unit controls. Output controls None. T ACC Decision rules NIA. Performanceindicators Broad and untested performance indicators in draft Management Plan.

Table 3.3. A summary of progress to date with respeet to data quality and research knowledge.

3.3.8 MONITORING, RESEARCH the usefulness of this approach has DIRECTION AND PRIORITIES yet to be demonstrated because catch a. The workshop considered it very rates may not be closely related to necessary to improve our stock abundance. Effort should be understanding of the relation standardised. This project should between fisheryperformance have a high priority. indicators and CPUE data. In the b. VMS data capture should be used to current management plan it is provide detailed stock assessment proposed to utilise catch rate to act information. With the introduction of as performanceindicators. However, VMS to the inshore trawl fleetthe

52 Scenario Risk Cost/Benefit • Recruitment protection 1. Growth overfishing. • Increased number and closures price of large prawns. 2. No growth overfishing • Loss of small prawns for no benefit. • Night-time area closure 1. Juveniles fished on • Loss of tiger and small within Moreton Bay movement out of bay. eastern king prawn (2-3 months) catch. • Reduced growth overfishing. • Extension and linkage 1. Insufficient protection • Loss of small eastern of two mile closures in of juveniles king prawn catch. areas where juveniles (within bay) • Reduced growth move out of bay. overfishina.

Table 3.4. Possible actions to be taken to the management of easternking prawns, their risks and benefits.

potential for its use to gather d. After much discussion on whether information on fleetdynamics and catch rates are representative of the the distribution of effort would be of fishery and good for performance tremendous value in monitoring the indicators, the consensus was that status of the stock. This will be there was no other likely approach difficult to developbut should have available at this stage. Assessing a high priority as if it proves juvenile easternking prawn catches successful could in the long run is not possible. The 'bay' prawn have a high returnfor a low cost. category used by fishers for any species of small prawn needs to be c. A fisheryindependent recruitment broken down to the species level. index from surveys is being This could be done by catch developed and could substantially sampling the 'bay' prawn to improve the sustainable management determinerelative proportions of of this resource. Tony Courtney is each prawn species at differenttimes undertaking this project, but some of the year. Another problem is that concernwas expressed as to the high eastern king and red-spot king coefficientof variationof his index. prawns are entered as one category It is recommended that a review of of 'king' prawns in the logbook. This the sampling intensity is made and,if could be solved by adding another more resources are not available, that column in the logbook or catch the time and spatial scale of the sampling in areas where both species project be redesigned e.g. only are known to occur. survey Moreton Bay during November.

53 3.5 REFERENCES

Barber, W.E. and Lee, C.P. 1975. Preliminary analysis of the physical factors influencingthe ingress of planktonic king prawn (Penaeus plebejus) post-larvae into Moreton Bay. pp.45-53. In: 'First Australian National Prawn Seminar'. (Ed. P.C. Young) (AustralianGovernment Publishing Service: Canberra) Coles, R.G. and Greenwood, J. G. 1983. Seasonal movement and size distribution of three commercially importantAustralian prawn species (Crustacea:Penaeidae) within an estuarine system. AustralianJournal of Marine and Freshwater Research 34: 712-43. Courtney, A. J.,Montg omery, S.S., Die, D. J., Andrew, N. L., Cosgrove,M. G. and Blount, C. 1995a. Maturationin the femaleeastern king prawn, Penaeus plebejus, fromcoastal waters of easternAustralia andconsiderations for quantifyingegg production in penaeid prawn populations. Marine Biology 122: 547-556. Courtney, A. J., Masel, J.M. and D. J. Die 1995b. Temporal and spatialpatterns in recruitment of three penaeid prawns in Moreton Bay, Queensland, Australia. Estuarine, Coastal and ShelfScience 41: 377-392. Courtney, A. J., Die, D. J. andMcGilvray, J. G. 1996. Lunarperiodicity in catch rate and reproductive condition of adult easternking prawns, Penaeus plebejus in coastal waters of southeast Queensland, Australia. Marine and Freshwater Research 47: 67-76. Courtney,A. J. 1997. Spawning stock dynamics of penaeid prawns in southeast Queensland and considerations forpreventing recruitment ovetfishing. Thesis submitted forthe degree of Doctor of Philosophy, University of Queensland, May 1997, 168p. Dakin, W. J. 1938. The habits and life-historyof a penaeid prawn (Penaeus plebejus Hesse). Proceedings of the Zoological Societyof London 108A: 163-83. Glaister, J.P., Lau, T., andMcDonall, V. C. 1987. Growth and migration of tagged easternAustralian king prawns, Penaeus plebejus Hess. Australian Journal of Marine and Freshwater Research 38: 225-42. Glaister, J.P.,Montgomery, S. S., andMcDonall, V. C. 1990. Yield-per-recruit analysisof eastern king prawns, Penaeus plebejus Hess, in easternAustralia. Australian Journal ofMarine Freshwater Research 41: 175-97. Gordon, G. N. G., Andrew, N. L., and Montgomery, S.S. 1995. Deterministic CompartmentalModel forthe EasternKing Prawn (Penaeus plebejus) fishery in New South Wales. Marine and Freshwater Research 46: 793-807. Grey, D. L., Dall, W., and Baker, A. 1983. A guide to the Australian Penaeid Prawns. pp.140 (Departmentof Primary Production, Northern Territory Government: Darwin). Kirkegaard, I., and Walker, R.H. 1970. Synopsis of biological data on eastern king prawn Penaeus plebejus Hess, 1865. Commonwealth Scientific and Industrial Research Organisation Division of Fisheriesand Oceanography FisheriesSynopsis No. 7.

54 Lavery, S. & Keenan, S. 1995. Genetic analysis of crustacean stock structure and stock size. pp.116-126 In: 'Proceedings of the workshop on spawning stock-recruitment relationships (SRRs)in Australiancrustacean fisheries. Joondoburri Conference Centre 1-3 June, 1994'. (Eds A. J. Courtney and M. G. Cosgrove.). (Department of Primary Industries, Queensland.) Lucas, C. 1974. Preliminary estimates of stocks of the king prawn, Penaeus plebejus, in South-EastQueensland. Australian Journalof Marine and Freshwater Research 25: 35--47. Montgomery,S. S. 1981. Tagging studies on juvenile eastern king prawns reveal record migration. Australian Fisheries 40(9): 13-14. Montgomery,S. S. 1990. Movements of juvenile eastern king prawns, Penaeus plebejus, andidentification of stocks along the east coast of Australia. Fisheries Research 9: 189-208. Mulley, J.C. and Latter, B.D.H.1981. Geographical differentiationof eastern Australian penaeid prawn populations. Australian Journalof Marine and Freshwater Research. 32: 889-895. Penn, J.W. 1984. The behaviour and catchability of some commercially exploited penaeids and theirrelationship to stock and recruitment. In: Penaeid Shrimps their Biology and Management. Gulland, J.A. and Rothschild,BJ. (Eds). Fishing News Books Ltd. Surrey, England, pp. 173-186. Potter, M.A. 1975. Movements of the easternking prawn (Penaeus plebejus) in southernQueensland waters.In 'First AustralianNational Prawn Seminar'. (Ed. P.C. Young.) pp. 10-17. (Australian GovernmentPublishing Service: Canberra.) Racek,A. A. 1959. Prawn investigations in eastern Australia. New South Wales State Fisheries Research Bulletin No. 6: 1-57 pp. Rothlisberg, P. C., Church, J. A. andFandry, C. B. 1995. A mechanism fornear shore concentration and estuarine recruitment of post-larval Penaeus plebejus Hess (Decapoda, Penaeidae). Estuarine, Coastal and Shelf Science 40: 115-38. Ruello, N. V. 1975a. An historical review andannotated bib liography of prawns and the prawning industry in Australia. pp. 305--41. In: 'First Australian National Prawn Seminar'. (Ed. P. C. Young.). (Australian GovernmentPublishing Service: Canberra.) Ruello, N. V. 1975b. Geographical distribution, growth and breeding migration of the easternAustralian king prawn Penaeus plebejus Hess. Australian Journalof Marine and Freshwater Research 26: 343-54. Somers, I. F. 1975. Growth of eastern king prawns (Penaeus plebejus). pp.104-11 In: 'First Australian National Prawn Seminar'.(Ed. P. C. Young.). (Australian GovernmentPublishing Service: Canberra.). Young, P. C. 1978. Moreton Bay, Queensland. A nursery area for juvenile penaeid prawns. Australian Journalof Marine and Freshwater Research 29: 55-75. Young, P. C., and Carpenter,S. M. 1977. Recruitment of postlarval Penaeid prawns to nursery areas in Moreton Bay, Queensland. Australian Journalof Marine and Freshwater Research 28: 74

55 56 en.> � � n ;::-· m � c...... �

UI en. --J C') -·� n ;:: � > � 8 t::::i r � - � C') r ;::i: � ;::i: -- 1:::1 � ('I:, � 4.1. PARTICIPANTS

Dr Ian Brown (Fisheries Scientist, DPI, Deception Bay) Mr Rik Duckworth (Fisheries Scientist, NT Fisheries, Darwin) Dr Tony Courtney (Fisheries Scientist, DPI, Deception Bay) Ms Cathy Dichmont (Fisheries Scientist, DPI, Deception Bay) Mr Mike Dredge (Industry Manager, DPI, Deception Bay) Mr Peter Gaddes (Trawl fisher,Brisbane) Mr Richard Gilbert (Trawl fisher,Bundaberg) Dr Neil Gribble (Fisheries Scientist, DPI, Cairns) Dr Malcolm Haddon (Head, Department of Fisheries, AMC, Launceston) Mr Simon Hoyle (Fisheries Scientist, DPI, Deception Bay) Mr Vern Lee (Trawl fisherand processor, Tin Can Bay) Dr David Mayer (Biometrician, ARI, DPI, Yeerongpilly) Mr Michael O'Neill (Fisheries Scientist, DPI, Deception Bay) Dr Ian Poiner (Fisheries Scientist andTRAWL SRC Representative, CSIROMarine Research, Cleveland) Dr Andre Punt (Fisheries Scientist, CSIRO Marine Research,Hobart) Dr Ian Smith (Trawl Managerand TRAWL SRC Representative) Mr David Sterling (Trawl fisher,Br isbane) Mr Clive Turnbull (Fisheries Scientist, DPI, Cairns) Dr Reg Watson (Fisheries Scientist, WA Fisheries, Perth) Ms Kate Yeomans (Fisheries Scientist, DPI, Deception Bay)

(No Trawl MAC Chair was allocated at the time of the workshop due to a secondment)

58 4.2. INTRODUCTION 1981). It is also assumed that the bulk of the fisherycatch and the spawning The fisheryfor saucer scallops, biomass relies on a single year class Amusium japonicum balloti, is an based on the recruits into the fishery importantcomponent of a multi-species that survives into their firstyear. trawl fishery on the East Coast of Queensland. Annual landings average Queensland's saucer scallopstock was about 1 200 tonnes of adductor meat, first fished in the mid-1950s, when with a landed value in excess of $25 m prawn trawlers working out of Hervey (Williams 1997). The scallop fishery Bay took appreciable quantities (Ruello takes place mainly between 21 °S and 1975). While annual landings have not 27 °S, in depths ranging from20 to 60 shown the spectacular variation often metres. It is regulated through input associated with scallop fisheries controls, which include limited entry (Hancock 1979), catch rates declined (which applies to the entire Queensland by an order of magnitude in the period east coast trawl fishery) and size limits 1980 - 1988 (Dredge 1988) and designed to optimise yield per recruit declined further in the mid 1990s (Dredge 1994). The fishery was (Williams 1997). The fisheryis characterisedby 24 hour fishing seasonal. Maximum catches and catch operations until 1988, but was limited rates occur in early summer months, to night-time only operations thereafter. when young of year scallops recruit into Three 10-minute by 10-minute areas the fisheryand adductor condition are were closed to trawling so as to act as at its peak(Williams and Dredge 1981). broodstock reserves in 1989, but were Variablesize limits apply to the fishery. repealed 15 months later due to policing A minimum size of 95mm shell length difficulties. Similar closures were applies during winter and is decreased again introduced in 1997 as a response to 90mm during the summer months to seriousdeclines in catch rates which (1 November to 1 April). This has the were observed in late 1996. These effectof amplifying the early summer closures were re-gazetted in early 1999. effortpulse (Dredge 1994). Saucer scallops have been shown to Average catch rates observed in late spawn in winter and spring, coinciding 1996 and early 1997 were less than half with water temperature change. It is of the 1988-1995 average for that time probable that saucer scallops are serial of theyear. This decrease in catch rates spawners, with females spawning more was of sufficientconcern to managers than once in a season (Dredge 1988). and fishers that emergency broodstock Growth is rapid, with animals attaining closures were gazetted. Resources were sexualmaturity at a shell height of 90 allocatedfor a large-scale survey mm or larger towards the end of their designed to establish a baseline firstyear of life(Williams and Dredge recruitment index of the saucer scallop 1981; Dredge 1981). Natural mortality resource. Additionally, data on scallop rates of adults are high, between 0.020 densities, size composition and 1 and 0.025 week- , suggesting that few distribution information was obtained. saucer scallops survive more than 3 By the time of this workshop only one years (Dredge, 1985a; Heald & Caputi of these surveys have been completed.

59 4.3. SAUCER SCALLOPS over all trawl sites within the three WORKSHOP GROUP preservation zones were calculated from the scallop data on size TASKS: composition fromthe October 1997 1. Value and cost of present survey. Since the length frequency of preservation closures. the total numbers in the preservation areas was required, the above measured 2. Value and cost of a possible frequencies were scaled to the winter/spring closure. proportion measured relative to total 3. Standardisation of effort. numbers caught. Total length frequencyover the whole zone would 4. Stock assessment model. therefore be calculated as: 4.3.1. VALUE AND COST OF PRESERVATION ZONES A101a1 C r r = l, NT r r : 4.1 This group was asked to investigate the 1,0ctobe , qA sampled long-termbenefits of the preservation r zones as well as the possibility of open ing these preservation zones annually where Nroc, , , is the total number of obe , for a short period to remove old, scallops in size class l, in October senescent animals. The preservation 1997 in preservation zone, r, zones have been introduced as spawn A 0101 A ampled 2 ing stock protection areas. The fishing ; and : arethe areas(m ) industry has argued that, given the trawled (sampled) and the total area mortality rate of these animals, few of (total) of preservation zone, r, the adults that have spawned in one C1,r is the number of scallops caught year would live to spawn in the next. in size class 1, in preservation zone Inorder to preserve these animals r, and subsequent to the spawning period q is the constant of proportionality, would therefore constitute a financial or catchability coefficient, estimated loss to the industry without benefiting at 0.5 fromLeslie-Delury egg and larval production. Thisgroup experiments in Western Australia therefore investigated the value of the using similargear (Joll & Penn, 1990) legal catch in the preservation zone at the time of the October 1997 survey Total meat weight (kg) was therefore and subsequently, also at the end of calculated frommonth-specific length December as December was the weight conversions using Williams and proposed monthfor opening the zones. Dredge(1981): The October survey had sampled sites at intensities somewhat higherthan the -5 . 23£-613.154 main fishinggrounds and all scallops WT NT r l,Octobe takenwere measured to the nearest r,r - lOOO l,Octobe ,r millimetre. : 4.2 The length-frequency(in 10 mm The value($) of this meat weight per classes, taking the mid-length of the length class and region was calculated class intervalas the actual length of using Table 4.1. individuals within that class) summed

60 Mid-length October December Numbers in each length class in each (I) Price($) Price($) preservation zone in December are 5 0 0 thereforecalculated by: 15 0 0 25 0 0 - T = T 9M 35 0 0 Nl,December,r Nl,October,re : 4.5 45 0 0 55 0 0 where M is the intrinsic natural 65 0 0 1 75 0 15.00 mortality of 0.025 week- 85 15.00 15.00 (Dredge 1985a). 95 17.00 17.00 105 18.00 18.00 As a result, the December meat value 115 18.50 18.50 foreach class can be calculated using data in Table 4.1. The resultant value of Table 4.1. Value of meat weight($) definedfor the resource in Yeppoon, Bustard Head each length-class. This table was obtained from consultation with industry members within the and Hervey Bay was about $255 000, group. $2 115 000 and $434 000 respectively. However, if the 90 mm minimum legal The resultant value of the resource was size would apply, then the catch would about $158 000, $1 581 000 and $310 be worth about $132 000, $1 538 000 000 for Yeppoon, Bustard Head and and $274 000 for Yeppoon, Bustard Hervey Bay preservation zones Head and Hervey Bay respectively. The respectively. However, if the 90 mm total legal standing stock value minimum legal size were to apply, then increased from $1 683 000 at the time the catch would be worth about $114 of the October survey to $1 944 000 in 600, $1 330 000 and $238 000 for the beginning of December. This Y eppoon, BustardHead and Hervey means that thelegal standing stock Bay respectively. value at the time of the October survey was 86% of the standing stock value in A von Bertalanffygrowth function to December. convert length, l, to age, a, was used to calculate age structureof the October In this analysis, the cost to the fishery population in each class: of the preservation zones has been interpreted as the loss of the legal a = - loctober October,/ _J_K ln(l 4.3 standing stock value. Other costs have L_ J : not been included. The value of the where K, the intrinsic growth rate preservation zones has been introduced was 0.05 week1, and as a spawning stock refuge that will Loo, the asymptotic maximum maintain the fishery through the length was 105.5 mm (Dredge, maintenance of recruitment. The group 1985b). concluded that it would be very difficult to assess thedollar value of the zones Animals in each length class that over the long-term as they did not know survive to December, would therefore the relationship between the amount of be 9 weeks older and have grown in stock preserved and the effectthis has length fromthe von Bertalanffy on the number of recruits the next year. formula: However, they did propose that exposing the preservation zones to = _ -Kav,c,mber,I ) t L_ (1 e :4.4 fishingpressure from1 December where l' is the December length could produce a $2.4 million dollar of length class, l. benefitto the fishery provided all the

61 legal size scallop in the preservation to be a major problem (M. Dredge zones were captured. Since most of personal communication). The second, these animalswould not survive to potentially more serious problem for spawn next year, it was proposed by the fishery, is that chipping to the shell this group that the fisherybe allowed could artificiallydecrease their size to fish animals larger than 90 mm shell below the legal minimum, thereby length during December of each year in increasing the period in which they are the preservation zones. unavailable to the fishery. The group recommended that more The proposal being considered was that in-depth analysis of this work would there be a winter/spring closure from be beneficial. It would be necessary to 20 September to 1 November. The simulate: advantages include: • the effectsof handling induced • A cessation of chipping by the high mortality and shrinkage through levels of effort required to capture chipping during the catch process; the minimum legal size of 95 mm. • the possible level of recruitment This also implies there would be that might be supplied by the uninterrupted growth of animals and preservation zones, given various a consequent increase in meat yield; spawner-stock-recruitment • An improvement in the condition relationships; of the meat with a consequent • inter-year variability and the improvement in the price (winter replenishment potential of various meat tends to have the poorest and 'closure' areas; lowest value); • • variation in meat value and meat A financial gain fromhaving no quality through the months; fuelcosts incurredto take scallops in the closure; • changes in fishing pattern outside • closure areas. An increase in the number of animals greater than the90 mm A large-scale simulation exercise would minimum legalsize afterNovember 1. therefore be able to address these issues. The disadvantages would include: • a loss of revenue fromthe legal 4.3.2. VALUE OF A POTENTIAL sized scallops captured during the WINTER/SPRING CLOSURE proposed closed period; The issue under consideration by this • potential for harming the market group was that, while legally fishingfor share by failure to provide product scallops greater than 95mm (the legal during the closed period; size during the winter months) the fishery catches large amountsof • natural mortality during the closed undersized scallops that must be period would mean there would be returned to the water. This lead to some population decline. undersize animals being damaged by The problem becomes one analogous shell chipping during catch sorting, to an analysis of yield-per-recruit. The which has two negative implications. industry would foregosome earnings First, the handling andchipping may through the reduction in landings result in the mortality of some of these during the closed period. However, small scallops. This is not considered the potential exists for an increase in

62 overall earningsthrough an increase in constant from the von yield deriving fromscallop growth and Bertalanffy. the avoidance of high levels of shell chipping, and the increase in the value Once the starting position in terms of of the catch through the improvements relative numbers in each size class was in meat condition. determined, the population was projected forward, again both in terms To investigate whether the closure of numbers and size structure, using would be advantageous to the fishing standard equations forgrowth and the industry required the group to application of natural mortality rates determine the size distribution of and fishingmortality rates. To follow scallops at the start of the proposed changes in size structure required data closed season. This then had to be on both the growth characteristics and projected forwardthe necessary six also estimates of the probability of weeks first while omitting fishing being chipped and by how much. In mortality and the effects of chipping, consultation with industry members, second while including fishing as usual. rough estimates on the extent of By comparing the outcomes and value chipping fromtagging data were used of the potential yield in each of these in the analysis to demonstrate the cases the merit of the proposed closure method. Once the forward projection would be indicated. was completed the relative numbers in each size class were translated into The group used data fromthe October meat weight and their relative value 1997 scallop survey within the determined and summed. Comparing preservation areas, which provided data the relative value fromthe population fromapproximately the middle of under conditions of closure and without October. This information was used to the closure demonstrated that the total project the population backwards in value of the fisheryincreased if a terms of both relative numbers and winter closure was put in place. growth to calculate the number and sizes of animals available at the The analysis was carriedout for the beginning of the proposed closure. northern and the southern area To project numbers backwards over separately as the growth characteristics time required estimates of natural of the scallops differsbetween the two mortality (because the original numbers areas. Data used was that collected were frominside the preservation areas fromthe preservation areas so as to fishing mortality could be ignored). To have a sample of animals thathad not project the size structure backwards been affectedby fishing mortality. required growth data (L- and K) and Without the existence of the the repeated application of Equation preservation areas this analysis could 4.6, which describes inverse von not have been produced. Bertalanffygrowth. The assumptions about chipping and its = prevalence mean that the particular L1+1 -L=(1-e-K) LI : 4.6 conclusions of the analysis are only e-K suggestive of the losses or benefits from the proposed winter closure. With the where Lt is the length of an animal approximate calculations the analysis at time t, suggested that by closing the fishery for L"' and K are the asymptotic those 6 winter-spring weeks, the fishery constant, and the growth would be enhanced in value by

63 approximately $1 million. This figure fisheryindicated that when GPS and ignores the savings introduced through plotters are used concurrently, relative thefishery not incurring fuel and crew fishing power increased by 7% over expenses. This suggests that, as a boats without such equipment (Robins minimum, more detailed investigation et al. 1996). This technology creep will of shell chipping should be made to also apply to the scallop fishery and remove the crude approximations therefore affectthe CPUE data. For assumed by this initial analysis.The this effect to be quantified, vessel financialbenefits to the fishery are, characteristic information is required, potentially, considerable. including when certain technologies were incorporatedinto the fleet In This group concluded that a late winter thisstudy, vessel informationwas not spring closure appeared to enhance the available and thereforea number value of the fishery by approximately unique to each vessel ('vessel sequence $1 million. It was recommended that number') was used as a surrogate. the real value to the fishery be determined. To do this the impact of A General Linear Modelling approach repeatedly landing the same undersized was applied to daily catch and effort animals on their extent and degree of records. Daily catch and effort data shell chipping should be investigated. from1988 to 1997 was extracted Also, the relative value andcondition covering the area from25 ° to 22.5°S. of animals taken in September and Factors that were considered to affect October, relative to those takenin catch rate (baskets.daf 1) were year, November should be formalised. Care month, vessel-sequence-number and must be taken to translate the research year-month interaction. The preliminary cruise measurements into those used by result is presented in Figure 4.1 below the commercial fishery.In the and shows that the real decline in recent workshop, a rapid set of measurements yearsis greater than the unstandardised provided an approximation of the data exhibits. The decline in relationship between the research and standardised catch rates over time is the commercial data. not significantlydifferent from zero. Furthermore, several highsand lows are 4.3.3. EFFORT STANDARDISATION not emphasised in the standardised data. Effectivefishing effort continually It should be pointed out that this increases in most fisheries, even though analysis is severely restricted by the the number of licence holders or total lack of specific vessel characteristic number of days fished each year may data and it was recommended that this remain constant. This continual 'effort information be obtained. creep' is characteristic of trawl fleets and is due to fishers adopting technological improvements in fishing practices, such as GPS andplotters. A recent study of the northernprawn

64 Year

OCl CJ) 0 .,... C\I (0 OCl CJ) CJ) CJ) CJ) mCJ) CJ)..... CJ) CJ) .....CJ) ...... CJ) CJ) .,...... 1.6

1.4 w 1.2 :::, a.. 1 0 Cl) 0.8

0.6

0.4 Standardised 0.2 i ---- i � Unstandardised 0

Figure 4.1. Unstandardised and standardised relative catch rates of scallop data from 1988 to 1997. Factors considered were year, vessel sequence number, month and year-month as an interaction.

Due to the size of the database and the by QFISH Grids). This largermodel is restricted time, a larger, more detailed more likely to capture the complexities model was only proposed. For this of the catch rate data. model, it was recommended that the fishingyear should start in November Although most of the daily catches since this is the month at which the recorded are:5 50 baskets (a 'basket' legal minimum size decreases to 90 mm contains about 500-600 scallops which shell height. Total effort in a month is 3-7kg of meat depending on season should also be included as it was and meat condition), some large daily considered possible that fishingeffort catch values were noted. These should becomes less efficient as competition be investigated (Figure 4.2). It may be for the resource increases i.e. catch necessary to remove some of these sharing. The year-monthinteraction extremes, as they may actually be data should be replaced with month as recorded in kilograms rather than a factor nested in year. Vessel baskets. Due to historical problems in characteristicdata, especially OPS the CFISH database, all catch data with plotters (guessed as a 10-20% entered into the 'weight' column of the effectover the total period), try gear CFISH system have been combined (:525% over the period), engine power with those in the 'number' column and (about 20% difference)and net interpreted as being in the units basket. size/configurationchanges (which At present, we are unable to discern could have affectedthe catch rate by between records that may be in 25%, but would have been a factorprior kilograms or in baskets, although the to this database in the late 1970s ), majority of the data are most likely should be considered. Further factors to be baskets. This problem was that could have affected the catch rates highlighted to QFMA prior to are responses to management and the workshop. spatial effects(which could be captured

65 16000 14000 0 C/J LO >. 12000 ,....CD � (.) C: 10000 /\ � Cl> � ..0 :::, 0" 8000 (.) u.e 6000 ,-... 4000 2000 0 0 0 0 0 0 0 0 0 0 0 0 0 0 C\J 'd" 0 -r- C\J M 'd" LO co ,-... co CD 0 -r- C\J M 'd" 0 -r- -r- -r- -r- -r- Catch (Basket)

Figure 4.2. Frequency distribution of daily catches in baskets. This data is a combination of catch in the CFISH 'weight' and 'number' column. Catches greater than 150 are more likely to be incorrectly assumed to be in baskets. Note that there are 347 records with catches greater than 150 'baskets', with two records being greater than 1000 'baskets'.

4.3.4. STOCK ASSESSMENT This group had to provide an estimate of biomass and yearly recruitment given available data of commercial catch and effort, the October 1997 survey estimate of total numbers, and tagging informationfrom the same survey. Catch was converted frombaskets to numbers using 500 as the conversion factor (Dredge personal communication.). An age-based model was developed keeping track of 1,2, ..,12 month old animals and thereafter older animals fall into a 1 plus group (a=13). The latter group was seen as being fullyrecruited to the fishery. The fullmodel fitseighteen parameters, minimising the negative log-likelihood:

: 4.7 where each of the components are described below.

4.3.4.a Initialnumbers Initial numbers for ages 1 in January 1988 is:

Nl988,1,l = (p IR1988 : 4.8 where N19s8,1,1 is the total population in year 1988 for the month January of age 1,

Numbers for ages 2 to 12 were calculated as follows:

h m. -M(a-l) N19 88,I,a = R1987w me : 4.9

where N19s8,1,a are the total population numbers in year1988 for January age a, R.1987 is the estimated recruitment for year 1987,

N19ss,1,13 = Nl : 4.10 where Nl is the estimated numbers of age 1+ in 1988.

4.3.4.b Dynamic processes

Annual fishing mortality, Fy,m was calculated as:

C F y,m : 4.12 y m = , "S� m,a N y,m,a

where Cy,m is the catch for year, y, and month, m, takenfrom CFISH, and Sm,a is the knife-edge selectivity by age andmonth, which includes the change in minimum legal size over the year.

= If Ism,aNy,m,a < Cy,m, then Fy,m 0.999 andadded to the likelihood as:

: 4.13

67 4.3.4.c Recruitmentpattern

r [-(111-111ea11_r)2slope_ l m _ (2var_r) J ',\.J'm -e : 4.14

where mean_r, slope_r and var_r are recruitment parameters to be estimated.

The resultant recruitment pattern was normalised to 1.

4.3.4.d Catch per unit effort

Mid-month numbers, Py,rn, were calculated as:

: 4.15

Estimated catch-per-unit effort is calculated as:

(;] = : 4.16 ln(qNov _ rPy,m) y,m where q is the estimated catchability coefficient, Nov_r is the November (only) adjusted catchability coefficient.

The negative log-likelihood function forthe catch and effortdata, assuming CPUE is log-normallydistributed, thereforebecomes:

n Le = nLn(fiiict)+ : 4.17 2

where n is 12 months*lO years and 6'is calculated as:

: 4.18 where q is the estimated catchability coefficient.

4.3.4.e Independent survey An independent survey was undertaken in October 1997, with a resultant absolute index of abundance and confidence intervals. The constant of proportionality of the survey was estimated by Leslie-DeLury trials using similar gear in WesternAustralia (Joll & Penn, 1990). As a result, the numbers fit was estimated within rangeof this value by adding to the likelihood functioni.e.

68 / / 13 ' ,2 1 survey,obs (N ) b n £.iN 1997,10,a - n 1997,10,adults N 1 ""' { (N1 991,10,s N1991,10,6 ) - ( :;';,�{o�j:v,naes _,____._a_=_1___ _,______.....,_ Ls \ln + ln )\2� ;urvey, juveniles + ;,,rvey,adults 20' 20' : 4.19

N:;;,:{ �; :veniles where 0 are the 'observed' juvenile (0-year olds) numbers from the survey, N:;;,']��;;1111, are the 'observed' adult (1+ year olds) numbers, and cr;,,rvey,adults cr;,,rvey,juveniles and are the survey variance forthe adult and juvenile estimate respectively.

4.3.4.f Tag model

Tagged scallops were measured and released during the October 1997 survey and the return data was availableth for analysis. Tag animals were assigned an ages 1 to 12 1 months and a 13 plus group (i.e. more than"" one years old) based on a von Bertalanffy growth curve with parameters L , Kand to of 105.5 mm, 0.05 week- and 0 years respectively (Dredge, 1985b). Numbers were updated per month in a similar manner to the dynamic model i.e.:

0for a=l �-M N�,m+!,a+l = sm,aFm,a { N�,rn,a(1 - for a+ 1 2 to 12 :4.20 2 s m �-M N: + N�,rn,13 (1 - m,13F ,13 for a =13

Estimated tag returns were calculated by: 1 N 1 F y,m £.i m,a y,m,a y,m C ="a S : 4.21

c;,m where N�,m.a and are the predicted numbers of tagged animals and the predicted numbers of tagged animals caught respectively.

The tag likelihood function (Li) therefore is describedbelow: i m ci,obs c ,abs I m m C�b £..£.i_ 'y,m y, y, c,o, st L = "y " m C + + ln( y,m J :4.22 where C�·.:' is the observed tag returnsin month, man year, y.

69 1400 ,------;::::======;-i --ObseMd 1200 -e- Predicted 1000

800

600

UJ 400 fs 200 0-+------�----1 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 Year

Figure 4.3. Model fit 'Predicted' to the 'Observed' catch rate data from 1988 to 1997 between latitudes 22.5 °S and 25°S from CFISH. Note that the annual peaks observed are generally predicted by the model, with the notable exception of 1997. The 1997 trends are influenced by an independent survey treated as an absolute index of abundance.

4.3.4.g Results intervals, whereas the early years are A fit to the observed catch rate data and estimated with less confidence. tagging data are given in Figure 4.3 and Narrower estimates will be possible if 4.4. Although the work is still at a very additional data, such as small-scale early stage, a surprisingly good fitto surveys completed in 1988-89, are the data was produced. Estimates of incorporated to this model. The group recruitment ± 1 standard deviation are suggested that a large-scale model be given in Figure 4.5. developed that incorporates size information and also is spatially These show that the survey data is explicit. instrumental in providing estimates of recruitment with narrowconfidence

70 • Observed ••• . . -• .. Predicted .- ,, ...... '\ . •' U) C: ::s 100 1D a: 50 O') ...... •... ea . -... ·•- ...... - .. I- 0 I'- I'- I'- OJ OJ OJ OJ O") O") 0) 0) 0) 0) 0) O") O") 0) 0) 0) O") 0) ,- ..... ,- ,- ,- ...... > (.) C ..c ...... t, 0 Q) cd Q) cd Cl. 0 z 0 ., LL � <( Date

Figure 4.4. Model fit, 'Predicted' to 'Observed' tag returnsdata. Tagged animals were released during the October 1997 independent survey over the whole survey area. Tags were returned by the fishing industry. Animals tagged in the preservation zones have not been included in this analysis.

3000 >< CU 2500 "C C: - 2000 C: CU E 1500 �::s ... 1000 a: 500

0 I'-

Figure 4.5. Model estimates of recruitment ±1 standard deviation. As expected, the largest uncertainty in the estimate is the 1987 initial recruitment estimate.

71 4.4. CONCLUSIONS and therefore a unique boat number, "vessel sequence number", was used 4.4.1. GROUP REPORTS as a surrogate. A peakin standardised catch rates in 1993 and thereaftera Preservation zones, that are closed to serious decline to historically low levels fishing, have been established in the in 1997 was evident. fishery as a spawning stock refuge. A model was developed to investigate A relative age model that utilises cost implicationsof these closures. unstandardised catch rate, independent Thisdemonstrated that opening these survey and tagging data was used to zones in December to expose animals estimate yearly recruitment levels. larger than90 mm (shell height) to The survey data was instrumental in fishingpressure could produce a providing recent year's estimates of maximum $2.4 million dollar benefitto recruitment with narrow confidence the fishery. The cost to the fisheryof intervals. It confirmsthe high recruit the preservation zones was interpreted ment in 1992 ( correspondingto a high as loss of the legal size standing stock catch rate in 1993) and a decline to value. historically low levels in 1996. The slight recovery of recruitment in 1997 A simulation model was also used to may be debateable, as the standardised test the benefitsand costs of a winter catch rate series does not confirm this closure from20 September to 1 upwardtrend in the unstandardised November. At present, animals larger data. than 95 mm shell height can legally be fishedduring that period. The model The independent survey was seen as considered shell chipping during the fundamental to the on-going sustainable catch process, growth rates, meat management of the resource. A clear condition and price per kilogram of decline in catch rates andestimated meat over time. This preliminary annual recruitment highlights the analysis concluded that a late possibility of unsustainablecatch winter/spring closure of the whole levels. Three preservation zones were fishery would enhance the value of the instituted in early1997 to address this fisheryby approximately $1 million situation. The independent surveys dollars. There are also implicit savings have demonstrated that these zones associated withsuch a closure e.g. in protect a significant proportion of the reduced operating costs. scallop population. At this stage it is not clear whether or how these areas, A generalised linearmodel was applied in fact,seed the remaining fishing to standardise the CFISH logbook catch grounds. rate data. This group highlighted problems with the catch data being 4.4.2. PROGRESS TO DA TE entered in the numbers and basket columns interchangeably. This A summary of progress to date in terms problem seems to be more chronic in of data quality and stock assessment the early years of the database. No knowledge is given in Table 4.3. vessel characteristic data was available

72 Category Comments Commercial Catch Potentially serious problems with the data as the catch has been entered in the basket or kilogram units columns interchangeable and incorrectly. Effort Fairly good, but unchecked data. Catch rate Historical data fromvoluntary logbook programs and other studies would extend the data to 1977, but is unavailable for analysis because the database entry is incomplete andunchecked. Data resides with QFMA. There is an urgent need to have the data developed and made available. Catch and effortseries for 1977 to 1980 is available. Recreational Catch, effort NI. and catch rate Independentindex of biomass or Large-scale independent survey since recruitment 1997. Estimates of natural mortality Good estimates of maximum natural mortality, but none are recent. Given changes in fishinglevels and present changes in benthic community structure, this should be re-estimated. Estimates of fishingmortality or Some estimates of F fromtagging data. biomass Relative age model developed in workshop, estimatesannual recruitment levels. Input controls Limited entry, gear restrictions, preservation zones, hull unit controls and size limits. Output controls None. TACC Decision rules NIA. Performanceindicators Broad and untested performance indicators in draft ManagementPlan.

Table 4.2. Summary of progress to date on data quality, stock assessment and management knowledge.

73 4.4.3. MONITORING, RESEARCH d. A stock assessment of thescallop DIRECTIONAND PRIORITIES resource has been shown to be a. Highest prioritywas given to the possible. This work needs to be scallop survey. The 1997 survey upgraded to a more complex model provided so much valuable data that that investigates the spatial aspects it ought to be continued, ideally of this fisheryas well as potential every year. It is, however, a costly future management options. Useful project andwill require financial extensions include; an improved support from theindustry. analysis of theeffects of handling induced mortality and shrinkage b. Improveour understanding of the through chipping duringthe catch relation between fishery process; determination of performanceindicators and CPUE recruitment levels by studying the data. In thecurrent management preservation zones (given various planit is proposed to put in place spawner stock-recruitment performanceindicators determined relationships); inter-annual by CPUE, but the usefulness of this variability and the replenishment approach has yet to be demonstrated potential of various 'closure' areas·' CPUEmay not be strongly variation in meat value andmeat related to stock abundance. This quality through the months, and; project should have a high priority changes in fishingpattern outside c. Use VMS data capture to provide closure areas. detailed stock assessment information.With the introduction e. Characterise the growth of scallops of VMS to the trawl fleetthe by region (latitude). In the potentialfor its use to gather workshop it was shown thatby informationon fleetdynamics and takingadvantage of closed areas at thedistribution of effortwould be the right time of year, millions of of tremendous value in monitoring dollars could be added to the value the status of the stock. This will be of thecatch. These analyses are difficult to develop but should have strongly influencedby the growth a high priority as, if it proves characteristics in each region successful, it could in the long run (which is known already to vary by have a high returnfor a low cost. A latitude). Characterisinggrowth by detailed study of the fleetdynamics region should thereforehave a high and effortdistribution in the scallop priority. Thisshould be tied in with fisheryand management systems completing the closurework done using seriallyclosed areas could be by Group 1 and 2 in this workshop used to optimise the value of the and may be captured by the model scallop catch. recommended above.

74 4.5. REFERENCES Dredge, M.C.L. 1981. Reproductive biology of the saucer scallopAmusium japonicum balloti (Bernardi) in Central Queensland waters. Australian Journal of Marine and Freshwater Research. 32: 775-787. Dredge, M.C.L. 1985a. Estimates of natural mortality and yield per recruit for Amusiumjaponicum balloti, as a function of exposure time. Journalof ShellfishResearch. 16(1): 63-66. Dredge, M.C.L. 1985b. Growth and mortality in an isolated bed of saucer scallops, Amusiumjaponicum balloti (Bernardi).Queensland Journalof Agricultural andAnimal Science. 42(1): 11-21. Dredge, M.C.L. 1988. Recruitment overfishing in a tropicalscallop fishery? Journal of Shellfish Research. 7(2): 233-239. Dredge, M.C.L. 1994. Modelling management measures in the Queensland scallop fishery. In: dredge, M.C.L., McLaughlin, R. and Zacharin, W. (eds) Proceedings of the Second Australian Scallop Workshop. Memoirs of the Queensland Museum 36(2): 277-282 Hancock, D.A. 1979 Population dynamics and management of shellfish stocks. In: Thomas, H.J. (Ed.) Population Assessments of Shellfish Stocks. Conseil internationalpour l'exploration de la mer, Rapports et Proces-Verbaux des reunions, Vol 175, 8-20. Heald, D.I. & Caputi, N., 1981. Some aspects of growth, recruitment and reproduction in the saucer scallopAmusium balloti (Bernardi)in Shark Bay, Western Australia. WesternAustralian Marine Research Laboratories Fisheries Research Bulletin. 25: 1-33. Joll, L.M. & Penn, J.W. 1990. The application of high-resolution systems to Leslie Delury depletion experiments for the measurement of trawl efficiencyunder open sea conditions. Fisheries Research 9: 41-55. Robins, C.M., Wang, Y-G. & Die, D.J. 1998. The impact of Global Positioning Systems and plotters on fishingpower in the Northern Prawn Fishery, Australia. Canadian Journalof Fisheries andAquatic Sciences 55(7): 1645-1651. Ruello, N. V. 1975. An historical review and annotated bibliography of prawns and the prawning industry in Australia. In'First Australian National PrawnSeminar'. (Ed. P. C. Young.) pp. 305-41. (Australian Government Publishing Service: Canberra.). Williams, L.E. (Ed.) 1997. Queensland Fisheries Resources: Current Condition and Recent Trends 1988-1995. Department of Primary Industries, Brisbane, pp 1-101. Williams, M.J. & Dredge, M.C.L. 1981. Growth of the saucer scallop, Amusium japonicum balloti Habe in Central EasternQueensland . Australian Journalof Marine and Freshwater Research. 32:657-666.

75 76 common name SEA MULLET

Mugil cephalus

77 5.1. PARTICIPANTS

Mr Jeff Blaney (Net Fisher, North Stradbroke Island) Dr Ian Brown (Fisheries Scientist, DPI, Deception Bay) l',,Ir Darren Cameron (East Coast Inshore Finfish Resource Manager, QFMA, Brisbane) Ms Cathy Dichmont (Fisheries Scientist, DPI, Deception Bay) Dr Malcolm Dunning (FisheriesScientist, DPI, Brisbane) Dr Neil Gribble (Fisheries Scientist, DPI, Deception Bay) Dr Malcolm Haddon (Head of Department, Fisheries, AMC, Launceston) Mr Ian Halliday (Fisheries Scientist, DPI, Deception Bay) Mr Simon Hoyle (Fisheries Scientist, DPI, Deception Bay) Ms Kath Kelly (Fisheries Scientist, DPI, Deception Bay) Dr Don Kinsey (QFMA Subtropical Finfish MAC Chair) Dr David Mayer (Biornetrician, ARI, DPI, Yeerongpilly) Mr Mark McLennan (Fisheries Scientist, DPI, Deception Bay) Mr Michael O'Neill (Fisheries Scientist, DPI, Deception Bay) Ms Michelle Sellin (Fisheries Scientist, DPI, Deception Bay) Mr Richard Steckis (Fisheries Scientist,WA Fisheries, Perth) Mr Wayne Sumpton (Fisheries Scientist, DPI, Deception Bay) Mr John Virgona (FisheriesScientist, NSW Fisheries, Sydney) Mr Lew Williams (Fisheries Scientist, DPI, Brisbane) Ms Kate Yeomans (Fisheries Scientist, DPI, Deception Bay)

78 Moreton Bay. The latter three account 5.2. INTRODUCTION: forabout half of the total annualcatch. SEA MULLET IN Excluding the periodfrom 1970 and 1987, the annualsea mullet catch from QUEENSLAND AND 1943 to 1995 has ranged from1 241 tin NEW SOUTH WALES 1960 to 2 686 tin 1988, with an average catch of 1 886 t (Figure 5.1). 5.2.1. FISHERY STATISTICS InNew South Wales, catches have The sea mullet (Mugil cephalus) is increased fromabout 3 000 to 4 000 t in an important commercial fisheries the last 15 years. While the estuarine resource in Queensland and New component of the catches has remained South Wales. at about 2 000 t, the ocean beach In Queensland, sea mullet sustains the component has increased fromabout largest single finfishfishery and arethe 500 t per annum to 2 000 t over the last mainstay of the freshfish trade. There 15 years (Figure 5.2). This component are1 039 licensed net fishers of the fisheryis comprisedlargely of in Queensland with 300 of these sea mullet migrating to spawn and has targeting sea mullet as one of their increased as a result of a growing principal species. Approximately 60 market forsea mullet roe. of these operate in Moreton Bay. The The key point about the ocean beach fisheryextends fromthe New South component of the sea mullet fishery Wales-Queensland border north to is that it targetsthe mature and pre Townsville. About 1 800 t is caught spawning population. We do not have annually between the border andthe sufficient informationto predict the northern tip of Fraser Island with a consequences of the recent increases further 200 t caught between Fraser in the oceaniccatch of sea mullet. Island and Townsville. The major sea However, management and industry mullet fishingareas areFraser Island, will become increasingly vulnerable the Sunshine Coast beaches, Moreton to the problems associated with over Island, North Stradbroke Island and exploitation if thisincreasing trend in

2800

2400

� 2000 0 1600

1200

O l 1, 1, 1, 1,,,,,,,,,, ·, ·,·, ·, ·, / :;:_,,4<--,-, --,,-,, -,-,.,., ...,.,,,r-r-, .,.,- 43 45 47 49 51 53 55 57 59 61Year 63 65 67 69 87 89 91 93 95

Figure 5.1. Total commercial catch of sea mullet in Queensland waters. Overall average catch is J 886 t. Note the breaks in both axes.

79 harvest continues without due regard valued at approximately 1.8 times for the conservation of the stock, and ($7.3 million) that of the estuary sector the maintenance of recruitment levels. and involves only half (380) the number of fishers. Almost half ($5.2 million) 5.2.2. VALUE OF THEFISHERY of the value of the total fisheryin New The Queensland fisheryis valued at South Wales is fromspawning run 7 to $8 M per annum. The estuarine catches on ocean beaches in the central catch is valued at -$2.3 M and the and mid north coast during April and ocean beach catch is valued at -$5 M. May. Half of all spawning fish caught for roe are takenby the 70 fishers licensed in 5.2.3. COMPONENTS OF THE the ocean beach fishery with theother QUEENSLAND FISHERY half caught by fishersin estuarineareas. The Ocean Beach Fishery is a While it once made up a substantial part specificallydesignated fishery, under of the fishery, the value of the 'hard differentmanagement from the gut' component has decreased to less remainder of the east coast net fisheries than $0.2M in the last 15-20 years. in Queensland. Itis a limited entry ('Hard-gut' is non reproductively fishery with 70 licences endorsed to active mullet that migrate during fishon the ocean beaches between the summer [Dec-Feb]) Queensland -New South Wales border and the northern tip of Fraser Island TheNew South Wales fisheryis valued fromthe 1st April - 30th August each at approximately $11.4 million per year. This fishery primarily targets the annum to the fishers,with spawning run of sea mullet. 70-80% approximately 870 fishersand about of the total sea mullet catch for Queens 380 taking90% of the catch. The land is takenduring the ocean beach estuarine sector of the fisheryis valued season from April to August (Figure at about $4.1 million and involves some 5.3) with 33% of the total catch taken 770 fishers. The ocean beach sector is during June. About half the catch

6000 -t!r-Total -U) 5000 -a-Estuary C C -o-Ocean t:. 4000

3000 t:,. 0 Cl 2000

1000

0 46 50 54 58 62 66 70 74 78 82 86 90 94 Year (ending June)

Figure 5.2. Commercial catches of sea mullet fromNew South Wales showing the ocean beach and estuarycatches from 1955-56 to 1995-96 and total catch from 1947-48 to 1995-96.

80 160 -••- Autumn __ ,._ __ Winter --•-- Spring - +- - Summer

a,140 G> ,, /'\, ,/ \ ' l '\.. I ' I '·, 2120 l \ l '\. I '- I '•. 0 \ �100 /I \ // '• .c ---· \ j -·· \ u ' ,/ -· is 800 ' ,,,,,. '.,-·· 'ii 600 400 +- _ � -._ ...... 200 ....__ --.,._____ - .,._...._ ---_ .... __..,. __ __ ..,_...__ _ - _...... :::: _-:::._r___ - _ ,... 0 ...... ------.------..----.....------...... -- ...... 1988 1989 1990 1991 1992 1993 1994 1995 1996 Year

Figure5.3. Seasonal changes of commercial sea mullet catches in Queensland. The dominance of winter and autumn is clear. Autumn is definedas March through May, Winter is June through August, Spring is September through November and Summer is December through February.

during the ocean beach season is taken otoliths (ear bones) in sea mullet are in estuaries,rivers and areas outside the formed annually. The ages of sea designated ocean beach fishery. Most mullet in commercial catches range of the latter is in spawning condition, from2 to 12 years and are dominated and is sold for roe at prices similar to by ages 2 to 6 years in Queensland and those receivedby the oceanbeach fishers. 3 to 7 in New South Wales. Estuary catches have higher proportions of The estuarine fisheryfor sea mullet younger fishand lower proportions of catches about 25% of the total sea older fishthan spawning run catches. mullet in landed in Queensland. On average the femaleswere larger Fishers in this fisheryoften operate as than males in the catches. This individuals or in pairsto take mullet differencein size is attributedmore in gill and tunnel nets. They provide to faster growth in females than a about 80 t of meat to local markets differencein the age structure of throughoutQueensland in all months catches between the sexes. other than in the spawning season. Catches fromestuaries are usually 5.2.5. SIZE AND AGE STRUCTURE lowest in spring, immediately after the winter spawning run, with increasing 5.2.5.a. Catch structure - estuaryand catches throughout summer and into ocean beach autumn (Figure 5.3). Intotal 10 432 sea mullet were measured fromthe 6 sites in AGEING 5.2.4. Queensland. Of these, 2 286 were Ageing studies indicate that the opaque returnedto thelaboratory forfurther rings observed in thin sections of the

81 15 a)All fish measured 15 b) 'Biological'Samples 35 c) Aged Sample 30 Estuary n 1257 12 12 .= 1007 25 1, OceanBeach n = 1007 I \ � 9 9 I .I' 20 I I \/1 I I CS I 15 I 6 \ I � I I 10 I 3 3 I ' ,, 5 \ \ \ \ 0 0 ... 0

200250 300 350 400 450 500 550 200 250 300 350 400 450 500 550 1 2 3 4 5 6 7 8 9 10111213 FL(mm) FL(mm) Age (Years) Figure 5.4. Size and age distribution of sea mullet fromQueensland estuaryand ocean beach catches. Data was derived froma) all fieldlength measurements, b) lengths offish fromsamples obtained from6 sites in Queensland and kept for biological analysis in the laboratory ('biological samples') and c) ages of these biological samples. Solid lines represent Estuarine fisherieswhicle dashed lines represent Ocean Beach fisheries. biological examination and 2 264 were in size from210 mmFL to 560 mmFL. successfully aged. Associated with the differences insize structure of the catch was a noticeable Sea mullet takenin estuarine catches differencein the age structure. Fifty were generally smaller thanthose taken three percent of the estuarine catch was in ocean beach catches (Figure 5.4a). less than4 yearsold while 25% of the The average size for estuarine caught ocean beach catch was less than 4 years mullet was 315 mmFL and for ocean old (Figure 5.4c). beach mullet 353 mmFL. The proportion of the catch over 370 mm 5.2.5.b. Estuarycatches FL in the estuarinecatch was 9% whereas in the ocean beach catches it The average size of fish caught in was 37%. Estuarinecaught sea mullet estuarieswas 313 mm FL in 1995 and ranged in size from190 to 490 mmFL. 316 in 1996 mmFL (Figure 5.5a). Ocean beach caught sea mullet ranged Females were larger than males with

--1995 ------1996 15 a) All Fish Measured 15 b) 'Biological' Samples = "fi12 1995 n 2166 12 1995 n 577 1995 n 577 1a 1996 n = 3337 1996 n = 680 1996 n = 680 (.) 9 9 0 .,,_a 6

3 3

200250300350400450500550 200250300350400450500550 1 2 3 4 5 6 7 8 9 10111213 FL (mm) FL (mm) Age (Years) Figure5.5. Size and age distribution of sea mullet caught in the Queensland estuaryfishery in two separate years. Data was derivedfrom: a) all length measurements; b) lengths of biological samples; c) ages of biological samples.

82 average sizes of 325 and 313 mm FL Estuarinecatches of fishers working in respectively. Juvenile fishcontri buted the Moreton Bay and Maroochy River about 10% of the catch for both years fisheries catch similar sized fish taken and were smaller thanmales and while the fish at theTin Can Bay site females with an average size of 279 were smaller. Average sizes for fish mm FL. Many of the juvenile mullet from each of the sites was 3 23 mm FL caught had reached the legal size and in Moreton Bay, 326 mm FL in the had entered the fishery, however many MaroochyRiver and 301 mm FL in were still sexually immature. Tin Can Bay. At the Moreton Bay and The age structure of theestuarine Maroochy River sites male and female caught fish was dominated by 2-7 year fishwere common between the ages of old fish,with femalesdominating the 2 and 6 years with juveniles less than 5 older age classes. There were also large years old making up about 7% of the differencesin age structure between totalcatch. The smaller size of the sea years. In 1995 catches were mainly mullet caught at Tin Can Bay is comprised of fishbetween 2 and 6 reflected in the age composition with years old with 5 year olds dominating the majority of the catch being less than the male catch and 2 and 5 year olds 4 years old. dominating the femalecatch. The 1996 catch was dominated by 2 year old fish, 5.2.5.c. Ocean beach catch newly recruited to the fisheryin that The average size of female fishwas year (Figure 5.5c). An interesting and between 45 and 50 mm larger than important featureto note about this the males at all sites with females differencein age structure between averaging 375 mm FL and males years is that the differencesare not 327 mm FL. reflected (Figure 5.5a) by the size distribution of the catch.

1995 ------1996

15 a) All fish measured 15 b) 'Biological' Samples 45 c) Aged Sample 40 1995 n = 2528 1995 n = 525 1995 n = 525 12 12 35 (.) 1996 n = 2401 1996 n = 482 1996 n = 482 -ea 9 9 30 (.) 25 -0 6 6 20 0 15 3 3 10 5 0 0 0 200 250 300 350 400 450 500 550 200 250 300 350 400 450 500 550 1 2 3 4 5 6 7 8 9 1011 1213 FL (mm) FL (mm) Age (Years) Figure 5.6. Size and age distribution of sea mullet caught in the Queensland ocean beach fishery in two separate years. Data was derived from:a) all length measurements; b) lengths of biolo{?icalsamples; c) a{?es of biolo{?icalsamples.

83 The age structure of the ocean beach Australia. Advanced stages of catch shows that fishfrom 2 to 8 years reproductive development in females old are commonly caught in the fishery. and males were foundat all sites in Fish over 6 years old were usually large Queensland and New South Wales, but females. The 1995 ocean beach season no runningripe females were foundin was dominated by 5 year old fish. The commercial catches. Therefore there is 1996 season was not dominated by any still no definite answer for the question particular age class, however, it is 'Where do mullet spawn?' interesting to note that about 17% of the catch in this year was of 2 year old fish, Gonad development of Queensland sea reflectingthe patternobserved in the mullet begins in autumn with increases estuarinefishery (Figure 5.6c). The evident in both male and female GSI 2 year oldscaught in the ocean beach by April. Maximum GSI (i.e. the fishery would have been newly proportion of body weight that is recruited fishthat had matured early, gonad) was attained in both sexes resulting in their participation in the between June and August with a return spawning run. Fish caught at Fraser to low values by October. Male GSI Island showed a greater proportion of during the peak spawning period 2 and 3 yearolds while 4, 5 and 6 year averaged 6.7% with a maximum of olds were dominant at the Sunshine 13.9% and femalesduring the same Coast and Stradbroke Island. Both periodaveraged 17.0% with a male and female sea mullet caught in maximum value of 24.6%. the ocean beach were represented in 5.2. 7. MOVEMENT older age classes than those caught in the estuarinefishery. A total of 2425 sea mullet were tagged and released in New South Wales 5.2.6. REPRODUCTION waters during 1995 and1996 (Figure The proportion of sea mullet involved 5.7). To the end of August 1997, 108 in spawning runs increases fromage (5.8%) recaptures had been reported out classes 3 to 6. It is likely that spawning of 1877 tagged and released in 1995 occurs in oceanic waters over a range and 59 (10.8%) out of 548 tagged and of latitudes along the East Coast of released in 1996 (Table 5.1).

Location Number Number Number Number tagged in recaptures tagged in recaptures 1995 reported by 1996 reported by August August 1997 1997 Tweed Heads (Hard-gut) 355 17 Tweed Heads (Spawning run) 653 22 277 25 Port Stephens (Spawning run) 287 23 162 8 Clarence River 495 39 72 25 Shoalhaven River 87 7 37 1 Total 1877 108 548 59

Table 5.1. Numbers of sea mullet tagged at various locations on the east coast of Australia between I Januaryand 31 June 1995 and the numbers ofreported recaptures up to August 1997.

84 ·, )fJ Fraser',,"·· Island Fraser Island Noosa River "-·-f:>,.--- . ',, Noosa River Noosa River - \, I \ � 1 Caloundra � .,,_ ; BRISBANE r ---, 2 ] BRISBANE .. '-\ ' f:-°3_:��::_-�_-.,,l·s Southport r ·i!/::"--,. l _ Currumbin Creek ·y/ Tweed River}::::-) \F) Tweed Rive :.:;�=:-/ · Tweed River )\}:;,- Clarence River ),��:)>t, Clarence River :.�: !_,_.�/" Richmond River ,.,�-"\\\<'\ \ Clarence River I\ \ \ I,\\ [ ) l / J Gotts Harbour Gotts Harbour I ; / I Ij 1 / /,/ / ;? / / - PortMacquarie PortMacquarie J - ,.' Port Macquarie / •. / \ y�, / / 1 )_;jv/ 1 _J.·-', Port Stephens ,,,;::;;,::. , / Port Stephens f ·- \2 Port Stephens Lake Macquarie "'::tf;J)JI/.;c::5,( ! ;/ / SYDNEY/.:_::-:_:, , ) SYDNEY -- -· _,, Hawksbury River Port Hacking Lake lllawarra ---··,.),,/ ''"'""'° """ ___r_J,J/ .,..,,= 1··--' "· Shoalhaven Bight / I """r Batemans Bay I lr ...."-. ;"\ \ )./ Nichols ,../ j �/ � _,r- Figure5.7. Map showing the movements of recaptured sea mullet in: a) a previous study ( Kesteven 1953 ); b) estuary releases; c) ocean beach releases in this study.

Tagging of sea mullet in New South 5.2.8. ASSESSMENT OF THE Wales during 1995 and 96 showed both LEGAL SIZE. northerly and southerly coastal Male sea mullet mature at about movements. Sea mullet tagged in 270 mm fork length or a total length estuaries were recaptured in or north, of 304 mm. Female sea mullet mature but not south, of the estuary of release at about 300 mm FL or a total length (Figure 5.7b). This is a similar result to of 338 mm. The differencein size at that in a previous sea mullet tagging first maturity is most likely due to study (Kesteven 1953) on the coast of differencesin growth rates between New South Wales. Jnthis previous male and femalemullet. study, the recorded coastal movements of sea mullet which were tagged mostly Of the 10 432 sea mullet examined in estuaries, were predominantly fromcommercial fishingoperat ions in northward(Figure 5.7a). Queensland 225 (2.2%) were below the legal size limit of 30 cm TL. Estuarine Sea mullet tagged during spawning runs fishing operations accounted for 197 on ocean beaches have been caught (1.9%) of these withocean beach north and south of the release site operations taking 28 (0.3%). In the during or after spawning migrations estuarine operations 74% of the (Figure 5.7c). There were many undersize mullet were taken during the examples of sea mullet tagged during summer period fromNovember to spawning runs and recaptured in the February. same or following spawning seasons. Several sea mullet tagged on Stockton Length measurements showed that Bight in the 1995 and 1996 spawning 1 691 (16%) fishwere below the size seasons were recaptured at the same at firstmaturity forfemales. Estuarine stretchof beach in the followingseason. catches contained 1 508 (89%) of these

85 while ocean beach catches contained 5.3.1. DISCUSSION OF EFFORT IN 183 (11% ). Assuming that 75% of the QUEENSLAND total catch is taken duringthe ocean This group mainly discussed the effects beach season and25% in the estuarine of fishbehaviour and logbook data fishery, with a 1: 1 sex ratio, 1.4% of quality on effort in Queensland. There total ocean beach catch and 3.4% of the are two main fisheries: an oceanbeach total estuarine catches would be of and an estuarine fishery. They immature females. thereforealso investigated a method by Under the current management which the effortof the different sectors arrangements used in the Queensland of the commercial fisherycan be commercial mullet fisheries, the legal separated, as gear type is not explicitly size limit and net restrictions are recorded in CFISH. This algorithm is allowing thevast majority of male and required, because the catch and effortof femalemullet to mature before entering these sectors represent differentfishing the fishery. practices. 5.3. MULLET TASK 5.3.1.a. Some factors affectingeffort GROUPS Mullet are highly migratory, aggregate and have a patchy distribution. This Due to limited available time, only means that effort is highly seasonal and three tasks were identified with rather that catch rates may not be indicative of broad portfolios: biomass. Furthermore, the use of the 1. Discussion on effort in terms of: logbook data for management as well as • Technological creep; stock assessment purposes, may mean that fisherswill misreport their effort • Integration with New and catch to protect their income South Wales; options. Since search effortis not • Definition of the two main entered into the logbook, particularly fisheries; with respect to the ocean beach fishery, • Seasonality of effort; effortwould remain underestimated. • Investigation into zero effort records especially search time; There was a suggestion that themarket • Location of effort. price of roe, rather than the availability 2. Examination of catch records: of fish, drives effort. Fishers areable to know the price beforegoing fishing and • Spatial distribution of catch; can avoid catching mullet if the price is • Fishery identification; too low. However, ocean beach fishers • Integration with New tend to fish irrespective of price during South Wales; the open season. • Catch rates versus year class strength. It should also be born in mind, that 3. Investigationof growth and ageing mullet can form part of a multi-species information: fishery. Certainly, the ocean beach • Comparegrowth data from fishery is generallylooking for any fish New South Wales, Queensland aggregations and not exclusively for and historical sources; mullet. • Produce a catch curve; In termsof effort creep, no detailed • Stock structure. vessel information was available. On discussion of the ocean beach fishery,

86 most of the technology and gear The followingspatial algorithm was changes that have occurred over the designed to identify the fisheryand history of the fisherytook place prior to gearused: the available effortdata (since 1988 Step 1. fromCFISH). Until 1986, there were • Link the sea mullet effort tofisher no limits on the number of operators records with a 'K' license (QFMA and only minor limits on equipment. Ocean Beach Licence) between the After 1986, restrictions on length and open ocean beach season of 1 April depth of nets were introduced. A limit to 31 August. This effort will be on the type of net materialhas not been classified as ocean beach effort. introduced and lighter synthetic gear has been in use before 1986. Inthe • All other records will be allocated 1960s, motorised tunnel net surfboats to the estuarine fishery. were introduced. At the same time, Step 2. tractors and trucks were used to haul • Since a tunnel net is legal only in nets onto the beach. In approximately Moreton Bay and Sandy Straits, a 1986, jet boats were introduced, which search restricted to these two areas, would have increased effort efficiency with the following rules,will by about 20%. The introduction of separate the tunnel and gill net good communication devices such as fishery. Net lengths of more than CB radios and cellulartelephones 800 and less than 2000 m will be occurred prior to 1986, but the defined as effort using a tunnel net. technology has improved over time. This improvement in communication • Net lengths less thanor equal to 800 was believed to have increased m, will be declared as gill net effort. efficiency by 4%. Step 3. • All remaining effort(i.e. in all other In the estuarine fishery, three main areas and net lengths less than 800 changes were discussed. Synthetic m), are to be declared 'other nets' nets were introduced prior to 1986, GPS have been used for tunnel netting 5.3.1.c. Recommendations in the 1990s and bycatch reduction This group produced an extensive list devices after 1996. None of these of recommendations, which have been changes were deemed to have had added to the finalmullet recommend a major effect on effort. ation list (numbers 4 to 11). 5.3.1.b. Algorithm to separate.fishery The estuarinefishery can be subdivided 5.3.2. EXAMINATIONOF CATCH RECORDS into a gill net and tunnel net fishery. This group investigated the spatial The gear used by these three fisheries distribution of the catch and the (ocean beach, estuarinegill net and effectiveness of the fisheryidenti estuarine tunnel net fishery) is ficationalgorithms developed by completely different and their effort the previous group. should thereforebe separated. The group had limited time to investigate This group attempted to identify the outliers in theCFISH database, but as differentfishery sectors using thecatch there were several such outliers data. They concluded that this encountered, they recommended these separation is not possible using the should be removed or investigated at catch data alone and that the method a future date. developed by theprevious group was

87 3500 ------� 3000 >- 2500 i 2000 - g 1500 - 1000 500

0 �t. ___..,.._...._...___._"T, -----.------,--+------! 0 20000 40000 60000 80000 100000 Daily Catch (kg) Figure 5.8. Frequency of mullet ( unspecified)and sea mullet catch weight per day from 1988 to 1997. the best method. However, they Rule 6: If catches are in Moreton Bay or modifiedthe previous groups Great Sandy Straits (CFISH grid algorithms into the following rules: numbers 'V34', 'W34' and 'W37') and (net length is longer Rule !:Delete all records where no net than 800 m and less than or length is recorded (about 9000 equal to 2000 m) then a tunnel records which is 10% of the net was used. total number of records), Rule 2: Delete all records where thenet Investigation ofthe catch rate length recorded is greater than frequency showed an extremely skewed distribution (Figure 5.8), with most of 2000m, 1 the catches being less than20 kg.day- • Rule 3:lf (net length� 2000 m) and The unstandardised catch rate corrected (not Ocean Beach 'K' licensed for gear type is shown in Figure 5.9. It fishers), then declare catch as is clear that the beach seine fishery's fromnon-ocean beach fishers catch rate is much higher than that of using gill nets. the othertwo fisheries. The gill and Rule 4: If(net length � 2000 m) and tunnel net catch rates have remained (Ocean Beach 'K' licensed fairly constant, whereas the ocean fishers out of ocean beach legal beach haul net catch rates varied 1 season) then declarecatch from between 400 and 1 400 kg.day - . Ocean Beach fishers using Investigation of the catch rates by gill nets. month clearly show that the tunnel net fishery catch rates peak prior to that of Rule 5:lf(net length� 2000 m) and the haul net fishery. (ocean beach 'K' licence fishers during ocean beach legal 5.3.2.a. Recommendations season) then declarecatch from Ocean beach fishersusing • It was the opinion of this group that haul nets, and the algorithm developed worked

88 1600

1400 --Gill --Haul 1200 --Tunnel ...... � 1000 G) 1i 800 .c 600 1i 0 400 200

0 CX) 0) 0 ,- (") LO (0 r--. CX) CX) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) ,- ,- ,- ,- ,- ,- ,- ,- ,- Year

Figure 5.9. Annual unstandardised catch rates separated by gear type. The separation was made using the decision rules described above.

well, but that the logbook system be 5.3.3. INVESTIGATION OF GROWTH so changed to rule out the need for AND AGEING INFORMATION algorithms in the future. During the introduction, it became clear Communication with fishersto that mullet are reasonably easy to age explain the importance of this filled using otoliths. It was therefore the in the logbook should also be group's task to definethe utility of this considered. About 10% of the ageing data forstock assessment records had no net length purposes andto consider whether the information and these catches were data would be of value for long term omitted from the analysis. monitoring. This group therefore • This group could not explain the investigated the age informationin high ocean beach haul net catch order to classify the age structure of the rates observed in 1992 and they resource, model a growth function and recommended that the data should estimate total mortality. be investigated further. Whether the 1992 catch rate high is real or a 5.3.3.a. Age structure data entry errorcould be determined Three gender categories were definedin by communicating with the three or the 1995/6-age database. These are fourmarketers forthis product. males, femalesand unidentified Furthermore, aninvestigation of (usually juveniles). Males and females apparent outliers may be directly generally follow the same trends in resolved with the fisherconcerned. relative abundance between age classes and year. The age data were also separated into fish caught by the ocean

89 beach and estuarine fisheries. In total Figures 5.10 and 5.11 highlight that the 1 257 and 1 007 otoliths were aged for catch-at-age over the long term may the beach and estuarinefishery show strong cohorts, which can be respectively. followedfrom one year to the next. The ocean beach fishery catch-at-age It is clear that the fisheriescatch a structure is clearer and covers a wider broad range of age classes, froma few range of size classes. Furthermore, 1+ year olds to 7+ and 8+ ages. In mullet are generally easy to age 1995, a large 5+ year class can be accurately and precisely. It would be observed in both fisheries (Figure 5.10). thereforepossible to model this This age class is still clearly visible, structure in order to estimate total especially in the ocean beach fishery, mortality, to follow resource age as a 6+ group in 1996 (Figure 5.11). structureand, with a longer dataset, be used in conjunction with catch rates for Virtual Population Analysis.

200 180 Estuarine: 1995 Cmale >,160 lilllfemale i•Total Cu 140 ::sCU 120 Cl)t:r100 LL 80 60 40 20 0 1 2 3 4 5 6 7 8 9 350 cmale 300 Estuarine: 1996 >, lilllfemale U 250 •Total C Cl) ::s 200

LL� 150 100

50 O, 1 2 3 4 5 6 7 8 9 Age Figure 5.10. Proportional age structure from the commercial estuarine fisheryfor the mullet (Mugil cephalus). The top panel isfor 1995 and the bottomfor 1996. Note the strong year class seen as 5+ in 1995 and 6+ in 1996.

90 5.3.3.b. Growth rates and estimates of of the von Bertalanffy growth natural mortality parameters are in Table 5.2. All theavailable fork length versus age data were used to estimate a von Parameter 1995 1996 Bertalanffy growth function parameters Linf 359.2 369.3 as well as obtain preliminary estimates K 0.458 0.504 1 of natural mortality (M yeaf ) (Figure Tzero -0.603 -0.369 5.12). These are data from commercial max Age 13 13 estuarine catches sampled in 1995 for males, femalesand unsexed, and Table 5.2 This group investigated the age juvenile data collected in 1995 at Tin information so as to classifythe age structure Can Bay and Maroochy. A least squares of the resource, model a growth function and minimisation procedure of residuals estimate total mortality. were utilised and the resultantestimates

450 400 Ocean: 1995 amale 350 ITfilfemal >,. 300 u •total C 1 Cl) 250 :::, i 0" 200 I Cl) I ... I u. 150 I 100 50 0 � j 1 2 3 4 5 6 7 8 9 140 Ocean: 1996 1 I 120 cmale lilllfemale' 100 ..,.j >,.u i I C I • total Cl) 80 I 0" Cl) 60 u. 40

0 1 2 3 4 5 6 7 8 9 Age

Figure 5.11. Proportional age structure fromthe commercial Ocean Beach fisheryfor the mullet (Mugil cephalus). The top panel is for 1995 and the bottom for 1996. Note the strong year class seen as 5+ in 1995 and 6+ in 1996.

91 500 X X X

X X X � � X �

400 X t X + + +H :f:!" + :f+ + 300 +\ ¥ 1 0 frf s::. ++ + .... + 0 + 0 I0 + &200 +

I + Estuarine j x Ocean Beach 100 o Juveniles

0+------,------,------.------,------,�-----1 0 2 4 6 8 10 12 Age Figure 5.12. Von Bertalanffygrowth model fitted to several datasources using a least squares approach. Data fromjuvenile and adult, males, females and unsexed fishwere pooled. Note that commercial fishers measure total length whereas fork length is used here.

Fork length varies enormously within a single age class, especially in the older and Pauly's estimator (Pauly 1984) on age classes. Furthermore, the female 1995 and 1996 data respectively. The length distributionsare much more natural mortality estimates were based varied than that of males. It would upon the growth parameter values from therefore not be possible to use length the 1995 and 1996 data analysis, in as a surrogate of age. which all the data collected in the relevant yearwas pooled. They The von Bertalanffy growth estimates included all juvenile data. These values of L,.,, K and t0 ranged from 340 to 380 1 are obviously very tenuous and no mm, 0.49 to 0.66 year- and -0.6 to -0.2 confidenceintervals were calculated on years respectively, depending on the any parameter estimates. combination of data included in the analysis. The commercialcatch data A simple model forestimating Z by may have been biased by gear followingthe 5+ to 6+ year class from selectivity, which was not considered. 1995 to 1996 was developed. Obviously, the dataset is too short fora Preliminary natural mortalityvalues of 1 robust measure. Also, emigration and 0.34, 0.51 and0.47 year- were derived immigration was not considered. The using the Hoenig et al. (1983) estimator results are given below (Table 5.3).

92 Estuarine Ocean Beach All Year-age m f Juv. Total m f Total Total 1995 - 5yr 92 82 3 177 111 100 211 388 1996-6yr 36.4 42.8 2.1 86.0 31.8 79.4 109.3 202.5 z 0.93 0.65 0.34 0.72 1.25 0.23 0.66 0.65

Table 5.3. Numbers of aged mullet in each age classes 5+ and 6+ collected in 1995 and 1996, separated by sex and fishery. Pooled frequenciesare also included forthe estuarine fishery, the Ocean 1 Beach fisheryand alldata. The resultant total mortality (Z year- ) estimate for each category is given. m are males, fare females and juv. are juveniles.

Total mortality(Z) was calculated as: 5.4. CONCLUSIONS

= Z ln Ns.1995 - ln N6,1996 5.4.1. GROUP REPORTS Catch and effortdata is not recorded by where Ns,1995 is the number of gear type other than by mesh net. The 5+ year olds in the sample. mesh fishery can apply three main gear types, being ocean beach seine, gill and The analysis of growth, natural tunnel nets. The catchability of sea mortality andtotal mortality is mullet for these gear types would be obviously extremely preliminary. different. Algorithmsbased on region, However, this group's work clearly month and licence package were shows the value of the age data. With developed to define catch and catch rate only two years of data, several methods records by gear type. In this analysis, of analysiswere achievable. Even several outliers were discovered. The though mullet grow to several years of catch rate frequency distribution was age, a long-term catch curve series was extremely skewed with most of the seen as being desirable in order to catches being less than 20 kg.daf 1. establish a good stock assessment The catch rates for the ocean beach model of the resource biomass. The seine fishery was much higherthan the animal is relatively easy to age anddata other two fisheries. The unstandardised is easily obtained. More information on gill and tunnel net catch rates remained juveniles would be needed than was fairlyconstant over the period 1988 to previously collected. The Ocean Beach 1997. Beach seine catch rates showed a data was seen as the most valuable large peak in 1992, however, the reality since it was less variable, gave greater of this peakwas disputed by some of age resolution between ages and had a the fishers present at the workshop. larger distributionof ages. However, if a model of mullet is to be seen as Age information was extremely useful capturing all aspects of the resource, it even though the dataset was short. The should include both New South Wales fisherycatches a broad range of age and Queensland data. Collaboration classes, from 1 + to 8+ year olds and between these states on this straddling therefore several years of data would be stock is therefore essential. required. The catch-at-age curve showed a strong cohort that could be followed over the two years sampled.

93 Total mortality values were estimated 5.4.2. PROGRESS TO DATE by following the fate of the strong A summary of progress to date with cohort. The analyses of this group was respect to data, stock assessment and extremely promising, but could only be managementare given in Table 5.5. treated as preliminary given the short dataset. 5.4.3. MONITORING, RESEARCH All available fork length versus age DIRECTION AND PRIORITIES information was combined to estimate 1. Age composition of the commercial von Bertalanffy parameters. Fork catch should continue. This length varies enormously within a appeared to produce the best index single age class and could not be used of the resource status. Due to the as a surrogate for age. Based on the many age classes, this will be a growth estimates, preliminary natural long-term venture in terms of mortality values were estimated. running Virtual Population Analyses, but the consensus was that the investment would be highly cost-effective and beneficial. A non-equilibrium yield-per-recruit

Category Comments Commercial Catch Spatial resolution not at Management Plan scale. Gear type below mesh net is not specified. Effort As above. Search time not recorded. Combine crews fromdifferent endorsements either log catch per endorsement (resulting in 2 days of effort being calculated) or as a single log (resulting in 1 day of effort being calculated). Definition of effortis unclear in multi-species and multi- endorsed fishery. Catch rate As above. Recreational Catch, effort Unknown, probably insignificant in and catch rate relation to commercial catch. Independent index of biomass or None. recruitment Estimates of natural mortality Preliminary estimates in workshop without variance estimation based on small dataset. Estimates of fishing mortality or As above. biomass Input controls Minimum legal size, spatial and seasonal closures on some sectors. Gear restrictions. Output controls None. TACC Decision rules NIA. Performance indicators None.

Table 5.4. A summary of progress to date in terms of data quality, stock assessment and management. 94 model could be attempted in the of gear used should be required. short term. A major deficiency is that until recently net type was not recorded, 2. Detailed ageing and growth studies so that mullet taken by ocean beach of O+ and 1+ animals are required to haul net could not be separated from corroborateexpectations concerning those taken in estuarine mesh and mullet stock productivity. Initially, tunnel nets. available data should be investigated in collaboration with 8. If possible, the catch and effortof New South Wales. the estuarinefishery should also be recorded at 6 by 6' grid. 3. Simulations are needed to investigate the Draft Management 9. In addition to sea mullet (Mugil Plan's managementrules in terms cephalus), several other mugilid of reference point trigger frequency species including tiger mullet (Liza argentea), mud or flat-tail 4. Greater spatial resolution is required mullet (L. subviridis), fantail mullet to fully analyse the ocean beach or flicker (Mugil georgii), andsand effort data. The standard 30 by or blue-tailed mullet (Valamugil 30-minute grid records are too buchaneni or V. seheli) - are taken coarse. The ocean beach fishery commercially along the Queensland should be required to log their data coast. These species represent less at a 6 by 6 minute (at least) scale. than 1 % of the annual mullet catch, This will allow greater spatial and the great majority is probably identification in terms of the taken in estuarine and tidal riverine legislated ocean beach zones in systems ratherthan on ocean the new Management Plan. beaches. However, there remains the possibility that some catches 5. A primary fisherfor multiple crews may have been mis-identified, or in the beach seine fishery should be that some records of (unspecified) nominated. At present, the catch mullet represent catches of species and effort of these logbook records other than sea mullet. would either be split between the licence holders or lumped into a 10. Investigate thepossibility of single catch record. This has a huge tracking the catch rates of a few effect on recorded effort,but with good fishersthat present good data, out a consistent direction of bias. as an alternativeor supplement to Any changes to the logbook should using all the fishers' data. take into consideration potential resource allocation problems. 11. Greater collaboration with New South Wales is needed, as this is a 6. Due to the multi-species nature of straddling stock. The New South the estuarine fishery, a target Wales catch seems to be higher than species recorded at start of fishing that in Queensland. New South day should be required. Gill nets Wales have recently attempted to are more specific interms of improve logbook information. targeting the catch than tunnel nets, They record their catch and effort since the mesh size allows for some monthly, with details such as the species selection. number of days fishedincludi ng search time, number of shots and 7. The catch by haul net should be catch by method. recorded i.e. greater differentiation

95 5.5. REFERENCES

Hoenig, J.M., Lawing, W.D. & Hoenig, N.A. 1983. Using mean age, mean length, and median lengthdata to estimate the total mortality rate. International Council for the Exploration of the Sea C.M. 23: 1-11. Kesteven, G.L. 1953. Further results of tagging of sea mullet, Mu.gilcephalus Linnaeus, on the easternAustralian coast. Australian Journalfor Marine and Freshwater Research. 4: 251-306. Pauly, D. 1984. Fish Population Dynamics in Tropical Waters: A Manual for Use with ProgrammableCalculators. ICLARM, Manilla.

96 common name TAILOR

Pomatomus saltatrix

97 6.1. PARTICIPANTS

Mr David Bateman (Recreational representative, SUNFISH, Brisbane) Mr Jeff Blaney (Net fisher, North Stradbroke Island) Dr Ian Brown (FisheriesScientist, DPI, Deception Bay) Mr Adam Butcher (Fisheries Scientist, DPI, Deception Bay) Mr Darren Cameron (East Coast InshoreFinfish Resource Manager, QFMA, Brisbane) Dr Tony Courtney (Fisheries Scientist, DPI, Deception Bay) Ms Cathy Dichmont (Fisheries Scientist, DPI, Deception Bay) Mr Mike Dredge (Industry Manager, DPI, Deception Bay) Dr Malcolm Dunning (Fisheries Scientist, DPI, Brisbane) Dr Neil Gribble (Fisheries Scientist, DPI, Cairns) Dr Malcolm Haddon (Head, Department of Fisheries, AMC, Launceston) Mr Ian Halliday (Fisheries Scientist, DPI, Deception Bay) Mr Jim Higgs (Programmer/Analyst of RFISH, QFMA, Brisbane) Mr Simon Hoyle (Fisheries Scientist, DPI, Deception Bay) Ms Kath Kelly (Fisheries Scientist, DPI, Deception Bay) Dr Don Kinsey (Subtropical FinfishMAC Chair) Dr David Mayer (Biometrician, ARI, Yeerongpilly) Mr Mark McLennan (Fisheries Scientist, DPI, Deception Bay) Mr Michael O'Neill (Fisheries Scientist, DPI, Deception Bay) Ms Michelle Sellin (Fisheries Scientist, DPI, Deception Bay) Mr Jonathan Staunton-Smith (Fisheries Scientist, DPI, Deception Bay) Mr Richard Steckis (Fisheries Scientist, WA Fisheries, Perth) Mr Wayne Sumpton (Fisheries Scientist, DPI, Deception Bay) Mr Lew Williams (Fisheries Scientist, DPI, Brisbane) Ms Kate Yeomans (Fisheries Scientist, DPI, Deception Bay)

98 6.2. INTRODUCTION year (Zeller et al. 1996). This is supported by monthly egg survey Tailor (Pomatomus saltatrix) are data from the Bribie Island area distributed worldwide, occurring in the (J. Staunton-Smith personal subtropical and temperate waters of communication) and data from Australia, Africa, the Mediterranean Miskiewicz et al. (1996). Eggs are and Black Seas, and the east coast of usually 0.9 - 1.2 mm in diameter. North and South America. In Australia, they occur between Fraser Island in Juveniles feed on small crustaceans, Queensland and Onslow in Western cephalopods and fish. Adults prey Australia. The east and west coast mainly on small schooling fishsuch as stocks are genetically distinct. Although pilchards, garfishand sea mullet. They they occur in southern waters, catches are cannibalistic and large individuals of tailor are rare in the Great Australian can be caught using tailor flesh baits. Bight, western Victoria and Tasmania. Because of this behaviour tailor schools tend to be of similar sized fish. The fish move northward to aggregate on ocean beaches in large schools Juvenile tailor enter estuaries at 35 to during late winter and spring. The 45 mm total length. At about 200mm largest known spawning aggregation FL juveniles move into open bay areas occurs at Fraser Island. Tailor are serial forminglarge schools. On reaching spawners, andrelease eggs and milt maturity, tailor move fromthe estuaries over extended periods of time. The onto the open surf beaches. Adult tailor eggs are pelagic and each female can may returnto estuarine and brackish produce up to one million eggs in a waters throughoutthe year. spawning season. Larvae hatch from eggs after about two days and remain 6.2.1. GROWTH in the plankton until they reach between Bade (1977) gives two alternative sets 35 and 45 mm. They then move into of von Bertalanffy growth parameters estuaries. (Table 6.1). He gave two options because he was uncertain how many Tailor are highly fecund, serial growth annuli were laid down each year spawners with an extended spawning on the scales he used forageing. period that may stretch out over many months. Monthly mean gonosomatic The FRDCIntegrated Stock indices indicatea peakof spawning Assessment and Monitoring (FRDC activity in southern Queensland T94/161) project's preliminary between September andOctober, with estimates of growth parameters are some spawning during the rest of the based on otoliths fromtailor frames

K Source L to Bade 1977 72.6 0.3267 0.296 Bade 1977 72.7 0.163 0.409 van der Elst 1976 75. l(FL) 0.197 0.0322 Barger 1990 101.9 0.096 -2.493 Barger 1990 94.4 0.180 -1.033 Terceiro and Ross 1993 94.6 0.242 -0.128

Table 6.1. A comparison of von Bertalanffy growth parameters from several publications.

99 collected from recreational fishers at samples come fromonly onshore Fraser Island during August 1995. stocks, andthere may be offshore Otoliths were aged whole under a light stocks where larger fish predominate, as microscope. These data come fromone occurs in Western Australia (Lenanton reader, and their precision and et al. 1996). If this is the case these reliability have not been estimated. mortality rates have been overestimated. The software package PCYield II (Punt 1992) was used to estimate parameters 6.2.2.a. Total Mortality (Table 6.2) of the von Bertalanffy growth curve, 1) Slope of Log(catch at age) against and to generate standard errors, given age (Butterworth et al. 1989). below. These results were L.,=40.0 1 ±2.28 cm, K=l.74 ±0.57 yr- , and 2) Z was estimated fromthe formula to=0.86 ±0.69 yr, based on a sample 1 size of 152. These results are unusual Z ln(l + with standard a a1 J, with a very low value of Looand very high k and to values relative to values 1 error (Chapman and Hobson fromboth the previous estimate from -JN the same population, and estimates 1960, Butterworth et al.. 1989). Age from other populations of the same at full recruitment (a1) was calculated species. The differencemay have been from the age of the youngest fully caused by the lack of fish older than 4 recruited age class in the sample. years in the 1995 sample, and problems Average age (a) included only fish with obtaining samples from enough older than a1. schools of tailor. The estimates of standard errorgiven Given these factors, it was decided by this methodare unrealistically to use one of Bade's estimates in the small, given annual variations in interim until further data have been recruitment, fishingmortality and collected and our results validated. Of naturalmortality. Data came from Bade's two parameter sets, we chose samples obtained on Fraser Island the one that gave the better least duringAugust 1995. As with the squares fit to our sampled data. This above estimator, the fact that only parameter set was Loo=72.6 cm, two fully recruited age classes were 1 K=0.3267 yr- , and t0=0.296 yr. observed makes this estimate unreliable. 6.2.2. MORTALITY Mortality was estimated in several 3) and 4) Length-based methods were ways under different assumption sets. used to estimate Z. Log(Number at Estimates at this stage suffer from the length)was regressed against log short time series of available data. They (1 - length/ Loo), the slope giving will improve as more years of data are Z/K (Galluci et al. 1995). Only collected. Methods forestimation of lengths fromfully recruited size total mortality (Z = fishingmortality classes were used. Length at full (F) + natural mortality(M)) are recruitment was estimated fromthe given below. mode of the length frequency distribution. It is ostensibly 27cm All of these methods assume that catch fork length (minimum legalsize), samples are unbiased with respect to but was observed to be 32 cm fork size and age. However the catch length from the length data. 100 Method for estimating Z Estimate + s.e. 1) Log catch at age against age 2.17 2) log(l + 1/(mean age - age at full recruitment)) 2.3 ± 0.26 3) Log(N at length) vs. log(l-L/L.,) - recreational data 2.19 ± 0.13 4) Log(N at length) vs. log(l-L/L.,) - commercial data 2.14 ± 0.37 5) Z=K((L., - mean length)/(mean length - lc,;t) - recreational data 2.7 6) Z=K((L., - mean length)/(mean length - Lc,;t) - commercial data 3.35 6) Mean of above estimates 2.48

Table 6.2. Total mortality (Z) estimated using four differentmethods, and the average value given by the methods.

The L= and K estimated by Bade 2) Hoenig (1983) equation was used to (1977) were used, together with the estimate natural mortality fromthe length distributions fromboth age of the oldest fish ever found,by commercial and recreational data. The the equation recreationaldata was obtained in both 1995 and 1996, and thecommercial M = el.44-0.982log(MaxAge). data came from 1996 only. Errorin the slope estimate was used to give a Max age was obtainedfrom the probability distributionfor Z oldest fishfound by Wilk (1977), in 5) and 6) Z was estimated fromthe a North American population of A ( saltatrix. The maximum age fromP. formula Z = K (Beverton Bade's (1977) study could not be : L -L Lent J used because he aged relatively few and Holt 1956), where Lcrir is the fish, and the population was heavily length at fullrecruitment, and r is exploited. mean length for fish greater than 3) The assumption made by Lcrit· Bade's (1977) von Bertalanffy Butterworth et al. (1989) forSouth parameters were used, together with African P. saltatrixwas used. length data fromthe August 1995 sampling program. 4) Rikhter and Efanov's (1977) equation: 6.2.2.b. Natural Mortality (Table 6.3) Natural mortality estimates were 1.521 M - - 0.155, similarly obtained in a number of ways. 07 1) Pauly's (1980) equation was used • tm to estimate natural mortality based where t is the age in years at which on von Bertalanffy growth m 50% of the stock is sexually mature. parameters and mean seawater Bade ( 1977) estimated that 'the temperature, according to the majority' of males were mature at formula: 26 cm and females at 28 cm. An -0.0066-0.28 log( 4.,)+0.651og( K)+0.46log( Temp) intermediate figureof 50% mature M = e at 26 cm was arbitrarily chosen. This correspondsto an age of 1.65. Von Bertalanffy growth parameters were obtained from Bade (1977). 5) The average of the above estimates.

101 Method for estimating M Estimate occurred adultfish move southward 1) Pauly 0.59 again. 2) Hoenig et al. 0.49 6.2.3. FISHERY 3) van der Elst 0.40 Commercialfishers use beach seine or 4) Rikhter and Efanov 1.16 w haul nets to catch tailor. This southern �ve�age�bove 0�66-� Queensland ocean beach fishery extends fromFraser Island to the New Table 6.3. Natural mortality (M) estimated South Wales border and focuses on using three different methods,and the average sandy beaches exposed to oceanic value given by the methods. conditions, particularly on the eastern shores of Stradbroke, Moreton, Bribie Fishing mortalitywas estimated by and Fraser Islands. However, this subtracting the average of the natural fishery primarilytargets the roe fishery mortality estimates from the average for migrating sea mullet (Mugil of the total mortality estimates, giving cephalus) on their seasonal pre F=l.8. spawning run up the coast from estuaries in New South Wales and It should be emphasised that estimates southern Queensland. Tailor are caught of total mortality (Z) also include in beach haul nets and may be taken as emigration fromthe fishery. If older age bycatch in mullet shots. However classes of tailor are less available to, or fisherstry to avoid such bycatch in less targeted by, the recreational and mullet nets because of the damage tailor commercial fisheries thanthe younger cando to the net by cutting the mesh age classes, then Z over-estimates total with their teeth. To minimise this mortality rate. This damage, tailor schools tend to be fished issue still needs to be investigated. with heavier ply nets. An essentially Tailor migrate northward duringlate small and incidental catch in the winter and spring. This movement is estuarine fishery is also observed. associated with the spawning run for Tailor are sometimes targeted with gill nets in estuaries, but this is not a these fishand often extends for hundreds of kilometres. The northern common practice. Recreational fishers tip of Fraser Islandhas been identified also catch migrating tailor. Both as a major spawning site for tailor with recreational and commercial sectors are large aggregations of ripe fish major stakeholders in this fishery. The appearing between August and October recreational catch of tailor in the each year. The spawning closure during southern region was believed to be at September between Indian Head and least as large as the commercial catch Waddy Point was introduced as a (Pollock 1980), catches fromFraser precautionary measure but because of a Island alone in 1979 amounting to number of uncertainties surroundingthe 180 tonnes. lifehistory of the species, it has not Both commercial and recreational been possible to evaluate the fisheriesoperate in a very narrow effectiveness of the closure. However, coastal band. Tailor only a fewhundred tailor egg distributions fromrecent meters off the coast are thereforeout of icthyoplankton surveys suggest reach to both sectors. Tailor in Western widespread spawning activity along the Australia are known to be distributed southerncoastline. Once spawning has both inshore and offshoreas far as the outer shelf (Lenanton et al. 1996). I 102 The commercial tailor fishery on the The commercial ocean beach fishery Queenslandeast coast reached a peak is restricted to 70 licence-holders in reported production of around400 Queensland and to a season extending tonnes during the 1960s, which lasted from 1 April to 30 August. Licence through to the mid 1970s. Since then holders may fish individually with a production has fallen to about 200 crew of 3 or 4 assistants. Operators tonnes per annum, but apparently for may also form groups in which the reasons other than overfishing. The catch is divided between several co market demand fortailor fell at about operating licence holders and their this time and has not returned to its fishery assistants. The amalgamation of earlierlevel. At the same time the level individual teams into groups ( which can of recreational fishingactivity increased comprise as many as 12-15 people) considerably, to theextent that now the results in a reduced level of competition recreational catch is estimated to be as and conflicton the beaches, and gives much as three to four times the size of all fishers concerned a fair share of the the commercial catch. Pollock (1979) total catch. estimated therecreational catch of tailor fromFr aser Island to be 180 tonnes, The commercial tailor fishery is driven while the commercial fishery in that by a variableand generally decreasing area caught about 25 tonnes. In 1995 marketdemand, but recreational there was a markedincrease in the pressure on the stock continues to recreational catch, attributedto a increase with the greater accessibility naturalincrease in the abundance of of remote beach areas resulting from tailor in inshore waters of northern the popularity of four-wheel-drive NSW and southern Queensland. vehicles. Of all of the State's marine finfish stocks, the tailor stock gives the Inthe commercial ocean beach fishery, greatest concernat present with respect tailor are takenexclusiv ely by haul or to potential overfishing. seine net. Occasionallyhaul nets are used in theestuarine fisherywhere the The commercial catch of tailor is structure of theshoreline permits (e.g. principally for a relatively small fresh around theRedcliffe Peninsula), but gill chilled market. No reliable figuresare (mesh) andtunnel nets arethe usual available forthe amenity value of the methods used in protected waters. recreationaltailor beach angling fishery, but the associated flow-on Gill nets capture fishby entanglement, to infrastructureindustries (beach usually by the gill covers or spines. vehicles, fishinggear, fuel etc.) would Mesh sizes of gill nets used for taking be substantial. mullet range from 3 to 3¾" depending on thetime of the year. The 3" mesh Commercial statistics relating to the effectively selects forlegal sized fish. tailor fishery is available for most of the period from1944 to the present, but the Tunnel nets areused to fencean reliability of the figuresis highly intertidal area thatis thenallowed to variable. During the post-war period drain off with theebbing tide. Mesh size until 1981 the Queensland Fish Board of the wings is usually 2¼" which is too (QFB) was the primary marketing small to gill most fish. The fishare agency for seafood products. The collected in a race staked below thelow Boardmaintained records of daily tide mark, from whichthey are sorted, landings (by species), but not fishing with undersized fishbeing effort. Illegal marketing (outside the immediately returned to thewater. QFB system) is known to have 103 450 400 � 350 - 300 250 200 'ii !:l 150 C: C: 100 50 0 1940 1950 1960 1970 1980 1990 2000 Year Figure 6.1. Commercial catch of tailor (Pomatomus saltatrix) in Queensland waters. occurred, but it is impossible to gauge The CFISH database does have the extent of this with any confidence. shortcomings. A major deficiency The landings figurestherefore under is that until recently net type was not estimate the actual landings by an recorded, so that tailor taken by ocean unknown and probably variable factor. beach haul net cannot be separated Between 1944 and 1969, landings were from those takenin estuarine mesh and recorded in pounds (lb.) whole weight. tunnel nets. The way the ocean beach From 1970 onward separate records fisheryfunctions also creates difficult were kept for whole fish (presumably ies with the available indices of fishing gilled and gutted) and fillets. Between effort. A large part of the effortin a 1970 and 1973 all records were haul net fishery for migratory schooling expressed as pounds, thereafter (from species consists of searching or 1974 onward) they were recorded as spotting, which is not incorporated in kilograms. For the purpose of our the compulsory logbooks. Another analysis, all figures have been complication is the ability of beach converted to whole weight (kg.) crews to join forces on an ad hoe basis, equivalent on the basis that 1 lb. = which clearlyinfluences effective 2.2 kg, and whole (gilled and gutted) effort. However, there is no way to weight = 2 x fillet wt. determine the size or number of crews DPI involved in any fishingoperation from In 1988 introduced a fishery-wide the existing dataset. compulsory commercial logbook program (CFISH). CFISH required The history of commercial tailor licence-holders to submit monthly catches in Queensland appears to catch returns detailing daily catch, comprise two phases (Figure 6.1). effort, and location. The logbook The firstphase, prior to 1975, is system was subsequently taken over by characterised by a mean annual catch of the new QFMA and became knownas 290 t, while from1976 onward annual the Queensland Fisheries Information catches have been consistently and System (QFISH). Unfortunately, substantially lower, at around170 t. for about seven years between the It is important to note that the sudden privatisation of the QFB and the decrease did not occur in the period of establishment of CFISH, no fishery change fromone reporting system to statistics were collected routinely the other, but while the QFB was in (i.e. apart fromshort-term voluntary operation. The sudden fall-offin research logs) in Queensland. annual catches of tailor may have been 104 related to marketdemand, which is According to the von Bertalanffy known to have slumped in the mid- growth curve, the growth in a year from l 970s when the QFB lost its regular an average length of 32 cm FL will be contract with the Queensland result in average lengthof 43.3 cm FL, Governmentfor the supply of fish or 48.5 cm TL. YPR at this length is products to the State's public hospitals 0.207 kg, an increase of only 3%. and institutions. Average Yield per Change The horizontal lines in Figure 6.1 are length at recruit (kg) ± from included to indicate the location of the recruitment standard current two long-termaverages rather than to to the error yield fishery suggest a lack of trend in the data time (TL in cm) series. After 1976, in fact, there 30 0.212 ± 5.5% appearsto have been a slight drop in the 0.002 size of the annual commercial tailor 35.8 0.201 ± 0.0% catch. To determine whether this 0.002 reflectsa similar trend in stock size 40 0.223 ± 11.0% would require a far more sensitive 0.003 measure of population abundance than 42.0 0.225 ± 11.8% (optimum) 0.004 raw annualreported catches. 48.5 0.207 ± 2.9% 0.004 6.2.4. YIELD-PER-RECRUIT Table 6.4. Yield per recruit at various YPR was modelled using the von average lengths (TL) at recruitment to the Bertalanffy growth curve from Bade fishery. The yield is the average yield per ( 1977) and the estimates of Z and M recruit when Fis 1.8. Natural mortality was described above. An iterative YPR set to 0.66 with standard deviation of 0.07. model was used (Hoyle and Sumpton The selectivity ogive was based on a CV of 0.15 with standard deviationof 0.03. in prep.) in order to estimate risk associated with the predictions, and to allow forerror in the input parameters. Ifthe minimum legal size were raised, The minimumlegal size ( with standard yield would not increase immediately. error)that would maximise the average There would be aninitial decrease in yield forF between 0.6 and 1.0 is 37.8 yield because of a relative lack of fish ± 2.3 cm FL, or 42.3 cm total length. in the older age classes. However, in the The current minimum legal size is long term ( assuming that the increased 30 cm total length. spawning biomass did not adversely affectthe number of recruits to the However, it is apparent fromthe length fishery)the average size of fishtaken data that have been collected that most would increase, with the total number fishdo not become available to the taken remaining below today's levels. Queensland tailor fishery until they Other factors to be taken into account reach approximately 32 cm fork length, are: or 36 cm TL. This gives an average yield per recruit of 0.201 kg. The optimum length of first capture is 42 cm TL, which would result in an average yield of 0.225 kg per recruit (an increase of 12%). However, achieving this length at first capture is not straightforward. 105 0.45 -.------,

.. 0.4 -+-1995 0.35 -1996 0 ----1997 C .3 0 t: 0.25 0 15" 0.2

C. 0.15

0.1

0.05

4 2 25 26 27 28 Latitude

Figure 6.2. Commercial tailor catch for years 1995, 1996 and 1997 by Latitude.

• the catch takenin New South 6.3.1. SIZE STRUCTURE AND Wales, DISTRIBUTION • the possibility thatolder fishmay At present, all size structure analysis differin their availability to the has been using samples unweighted by fishery, catch size. Samples have been collected over a wide range of latitudes. • the possibility that the different Workshop participants highlighted that sectors involved in the fishery may unweighted length-frequency have differingfishing objectives, distributions may be biased towards • and that increases in biomass of large catches froma specificlatitude. older (larger) fishmay have a density-dependent effecton the Commercial size frequency data was mortality of younger age-classes. available for 1996 and 1997 from the 'Integrated Stock Assessment and 6.3. TAILOR Monitoring Program' (FRDC T94/161), with the latter year having the most TASK GROUPS comprehensive dataset. The relevant year's commercial catches by latitude 1. Size structure and distribution. forthe months Aprilto August were 2. Estimating relative commercial and used to weight the size frequency recreational harvest. (Figure 6.2). Large relative catch 3. Analysis of recreational and changes by year and by latitude can be commercial catch rates. observed. In 1995 and 1997, most of the catch was taken fromlatitude 27°S 4. Investigation of ageing and growth ° ° estimates. (27 27.99 S), which corresponds to Moreton Bay and the offshoreislands.

106 0.12 1996 m:iWeighted 0.10 • Unweighted 0.08

0 0.06 0 0 0.04

0.02

0.00

0) 0 LO C\I I.C) (0 Fork Length Figure 6.3. Weighted and unweighted commercial tailor size-frequency distributionsfor 1996. Verylittle differencecan be observed between the differentsize-frequencies.

0.14 DWeighted 0.12 1997 • Unweighted 0.10 ....8 0.08 &_ 0.06

0.04

0.02

0.00

00 - N \C) Fork length (cm)

Figure 6.4. Weighted and unweighted tailor commercial size-frequency distributions. Only a slight change in distributionscan be observed.

107 Weighted and un-weighted commercial first time that size of the recreational size frequency distributions are given catch of tailor could be estimated. for 1996 and 1997 (Figure 6.3 and6.4 ). Although there is a slight shift to larger In the workshop, a firstattempt at a length frequencies for the 1997 data, no very rough mean and upper limit of the effectof weighting was observed in the recreational catch was estimated on the 1996 data. The lack of change in 1996 basis of the followingassumptions: was due to each of the three latitudinal • a logbook participation loss of 10% ranges, which reported catches, in the second quarter of the survey, effectively having identical catches. followed by 17 % in the third and This was not the case in 1997 and so 33% in the fourth quarter. the weighting had a discernibleeffect. • The 1996 data includes a sample in The survey covered 0.75% of the total recreational fishingcommunity which large tailor were targeted. The group recommended that this sample (i.e. 5 000 out of a possible 667 600 people). The maximum catch figure should be removed. assumes that 0.5% of the total Although the effect of weighting was recreational fisherspartic ipated not large, the group recommended that in the survey (i.e. 1 in 200 of the this procedure should be continued in 667 600 fishers). further analysis. Any major latitudinal • The average weight of a tailor was shift in catch would therefore be taken estimated as 0.4 kg from into consideration. This work will be recreational size frequency data. reported in the FRDC T94/161 'Integrated Stock Assessment and A firstestimate of recreational tailor Monitoring Program'. catch (tonnes) resulted in anaverage of 292t and a maximum of 437t. The 6.3.2. ESTIMATING RELATIVE commercial harvest based on CFISH COMMERCIAL AND RECREATIONAL is between lO0t to 200t per annum. HARVEST Obviously, this work is extremely Total recreational and commercial catch preliminary, but does show that the is not a good estimator of tailor stock recreational catch is of thesame health, but does give an index of the magnitude or greater than the RFISH relative resource share between the two commercial catch. The program major fishingsectors. Unfortunately,in data will be analysed formally by the the past, there has been no data Australian Bureau of Statistics and available as to the recreational catch. reported on by QFMA. Only the commercial catch is monitored 6.3.3. ANALYSIS OF RECREATIONAL in theQFMA logbook program. There AND COMMERCIAL CATCH RATES has, however, been much debate as to the recreational catch and its size A comparisonof commercial and relative to the commercial catch recreational catch rates has never because the resource is perceived to previously been completed using be under heavy fishing pressure. standardisation methods, although these QFMA has recently undertaken an datasets can not be combined as they extensive recreational telephone and cover different time periods etc. The logbook survey (RFISH). They have group's objective was to standardise the kindly made their preliminary data two datasets by year, month and other available for analysis. This was the factors, if necessary, and then to compare the relative trends. 108 Location Location clubs that contributed only a small Number amount of data. The remaining 30 Caloundra 40 Double Island Point 400 records were used. 41 Double Island Point Surf 62 Jumpinpin An analysis prior to the workshop 63 Jumpinpin Surf Bar indicated that the average weight of 80 Moreton Island tailor.fisher-1 .trip·1 had not changed 81 Combl Point 84 Reeders Point, Moreton Island significantlybetween 1973 and 1991. This work will be reported on in the 86 South Moreton Island 87 Moreton Island Surf 'Integrated Stock Assessment and 93 North Stradbroke Island Monitoring Program: report (FRDC 114 South Stradbroke Island 115 South Stradbroke Island Surf T94/161). A General LinearModel of 182 Surf Undefined the ocean beach recreational club catch rate data as the natural logarithm of Table 6.5. Locations and theircode within 1 1 the recreational club dataset used in the analysis number of fishcaug ht.fisher" .trip- + 1 of effort standardisation. was modelled, with year, month, club and location as factors. The location and year factors were significant. The resultant standardisedgraph shows a major increase in relative catch rate during the late 1960's and early1970's, 6.3.3.a. Recreational club data whereas, following a decline to 1960s levels, the catch rates have remained Recreational club data was kindly made relatively stable in the last decade available to the workshop by QFMA. (Figure 6.5). The standardisation had its The group selected recreational club largest effecton years where little data data from ocean beach areas only was available, tending to bring these (Table 6.5) so as to be able to compare yearsto a similar relative catch rate with commercial ocean beach catch as adjacent years. rates. The analysis discounted small

8-.------, Recreational Club (Ocean Beach) 7

K 5 0 Q)> 4 3 i"iii � 2� __.....,. _ __ ..,.

0 -----,----,----,.---,------,----,,--,----...-....----l 1937 1942 1947 1952 1957 1962 1967 1972 1977 1982 1987 1992 Vear

1 1 Figure 6.5. Standardised recreational club catch rate (number fish caught. fisher- .trip- ) data from ocean beach areas 109 6.3.3.b. Commercial catch rate data is unique to a vessel or seining Thetailor catch and effort data was operation with a specific set of allocated to the two fisheriescategories; characteristics. Ifany major change the ocean beach and an 'incidental' to the vessel or operation occurs, then fishery (which would be mostly a new unique number is given. estuarine caught fish). The algorithm A non-significant lineartrend (p=0.068) used to divide the catch and effort data in the ocean beach fishery catch rates was to definethe Ocean Beach catch as was observed (Figure 6.6). A similar tailor caught between April and August trend was observed with the 'K' QFMA by a licensed (Ocean Beach standardised 'estuarine' catch rate licence) fisherand theremaining weighted GLM (Figure 6.7). There records as 'incidental'. The 'incidental' is very little difference between fishery was not separated further. standardised and unstandardisedtrends. To remove outliers andpossible Standardisationof the 'incidental' incorrect allocation to fishery, records fisheryseems to have had a very small with catches less than 100kg forthe effect on the trend. Itwas suggested ocean beach fishery and catches greater that the most important information for than 100kg for the 'estuarine' fishery commercial effort standardisationis were eliminated from consideration. market price, since the behaviour of the A weighted (by number) GLM of the fisher is affectedby this factor. This natural logarithm of the two catch rates information was not available at the (plus one to avoid taking the log. of workshop and it was recommended that zero) with year, month and 'vessel a more in-depth analysis of this data sequence number' as a factor was with price be undertaken. undertaken. 'Vessel sequence number'

1.8 1.6 • w 1.4 • ::> 1.2 1 ,. - -· . -• - , , .. ·- 0.8 - --. � 'ii 0.6 • a: 0.4 ______Standardised ,,___ 0.2 - • - Unstandardised 0 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 Vear

Figure 6.6. The comm�rcial ocean beach standardised relative catch rate froma weighted (by number) General Linear Model (GLM). A non-significant(p==0.068) trend was observed

110 1.2 --- Standardised 1.15 •· Unstandardised • ::::, 1.1 • c. 1.05 (.) • /\. Q) 1 > .. . \, .. ;:; 0.95 \. 0.9 \ • 0.85 0.8 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 Year

Figure 6. 7. The commercial 'incidental' standardised relative catch rate from a weighted ( by number) GeneralLinear Model (GLM). A significant declineover years (p=0.007) can be observed.

0.7 ,------, � 0.6 C � 0.5 g 0.4 J: §! 0.3 i 0.2 � 0.1 0 -1--�-,---.....----,.---,..---.---1 0 2 3 4 5 Age (years)

Figure 6.8. Comparison of ageing frequenciesby two readers, Ian Brown ('IB')from DPI, Deception Bay and Richard Steckis ('RS')from WA Fisheries, Perth. Note that RS tends to age the same ototliths as being older than IB's ageing.

6.3.4. INVESTIGATION OF AGEING • bias i.e. assigning ages that are ANDGROWTH ESTIMATES younger or older; or There was great concernat the very • precision error i.e. high variation low proportion of older fishin the that has lead to an overestimate of commercial catch samples. If this total mortality. factor is a reflectionof reality, then this resource is being subjected to enormous 6.3.4.a. Bias fishingmortality. This group, therefore, In terms of bias, a scientist from had to address the question of whether WesternAustralia working on tailor, the ageing technique used by the FRDC RichardSteckis, was asked to age T94/161 IntegratedStock Assessment sectioned otoliths. Richard tended and Monitoring Project (ISAMP) team to age these otoliths older than those is subject to: 111 within the Centre (Figure 6.8). As a In this experiment different readers result, his estimates produced a lower within the Centre read whole and 1 total mortality estimate of 0.81 year- sectioned otoliths twice. In Table 6.6 1 as opposed to the 2.0 yeaf • However, an example of averaged double ageing the project had favoured whole otolith of whole sectioned otolithsfor readers readings and did not use sectioned Ian Brown and Michelle Sellin otoliths for most of their study. Wayne demonstrates that the level of Sumpton from DPI who was originally disagreement between two independent involved in this project was therefore readings by a single reader of 328 asked to age a few of the whole otoliths are high for tailor and that otoliths. His results tended to agree reader bias changes with age of fish. with the estimates from the ISAMP project. This highlighted that there Since this table can be interpreted both is much disagreement with the ageing along the row anddown a column, the between States and that validation average of these two approaches is required. resulted in the following normalised transition matrix (Table 6.7). For ages 6.3.4.b. Precision 4 and 6, no data was available and The IS AMPageing experiment's results thereforeno variance or bias was were used to adjust for the degree of assumed. variance in ageing of a single otolith.

Age 0 1 2 3 4 5 6 0 ! 0 0 0 0 0 0 1 1 190 12 0 0 0 0 2 0 17 96 4 0 0 0 3 0 1 5 Q 0 1 0 4 0 0 0 0 Q 0 0 5 0 0 0 0 0 .Q 0 6 0 0 0 0 0 0 0

Table 6.6. Comparison of average paired otolith readings of the same set of otoliths by two readers combined. Highlighted numbers show number of otoliths pairs which were aged the same. As fishage increases the bias changes.

Age 0 1 2 3 4 5 6 0 0.75 0.00 0.00 0.00 0.00 0.00 0.00 1 0.25 0.92 0.13 0.07 0.00 0.00 0.00 2 0.00 0.07 0.84 0.86 0.00 0.00 0.00 3 0.00 0.00 0.04 0.00 0.00 1.00 0.00 4 0.00 0.00 0.00 0.00 1.00 0.00 0.00 5 0.00 0.00 0.00 0.07 0.00 0.00 0.00 6 0.00 0.00 0.00 0.00 0.00 0.00 1.00

Table 6.7. Normalised transition matrix demonstratingbias and variance of paired otolith readings. For ages 4 and 6 no data were available.

112 1200 l •observed D'True' lOXJ I G° 800

600

400

200

0 1 2 4 0 3 5 6 Age

Figure 6.9. Dijference �etwe:n 'True' and the observed age struture. The 'true' age structure attempts to take mto consideration the bias and variance of paired otolith readings.

Using all the 1997 age data, it is 6.4. CONCLUSIONS possible to calculate the true age structure (Ntrue,a) by multiplying the 6.4.1. GROUP REPORTS inverse of the transition matrix There was great concernexpressed at (T a e.a e') with the observed age g g the very low proportion of older fish in structure (Nobserved,a e) i e g . .: the commercial catch samples. Ifthis 6 l factor reflects the true situation, this N observed,age =Tage,age Ntrue,age ·• • resource is being subjected to enormous fishingmortality that could not be The differencebetween these age sustainable. Two possible error sources structures (Figure 6.9) are not as could bias thecatch-at-age curve. marked as those between otoliths aged Otoliths may be incorrectly interpreted by scientists fromthe two different as being younger than they really are or Centres. large and older animals tend to remain 1 The resultant total mortality (year" ) offshore and are therefore not froma catch curve analysis changes vulnerable to the commercial and from 1.9 to 1.3 for theobserved and recreational fishery. This situation has 'true' respectively. Given this high been observed in Western Australia, total mortality, the resource would still but was thought to be unlikely in be interpreted as being overexploited. Queensland waters.A scientist from Itwould thereforebe fundamental to Western Australia, invited to participate the management of this fisheryto in the workshop, read andinterpreted validate the ageing of tailor or to clearly the otoliths of Queensland animals demonstrate that these animals cannot previously aged by Queensland be aged with high confidence. research staff. The bias and variance between the readers fromthe centre and Western Australia, within the centre and between subsequent readings of the same otolith by a single reader were analysed using a transition matrix modelling approach. Total mortality 113 estimates could range from 0.8 to 2 1 6.4.2. PROGRESS TO DATE year- depending on the interpretation A summary of data quality, stock of the ages. these values, whether assessment and management interpreted from the optimistic to the knowledge is given in Table 6.8. pessimistic, are still very high in relation to estimated natural mortality 6.4.3. MONITORJNG, RESEARCH rates. Ageing issues should be resolved DIRECTION AND PRIORITIES as a matter of urgency. If total mortal 1. Validation of the ageing of tailor is ity rates are found to be high then the essential to the future assessment costly exercise of determining the and management of this fishery. presence of offshore, large animals In the workshop it became apparent would be essential. that the health of the stock was either good or bad depending A first estimate of recreational tailor entirely upon whether the ageing catch fromthe recent QFMA studies already carried out are recreational survey showed the correct or biased downward. For a recreational catch to be about 290t small financialoutlay a large return (with maximum levels of about 440 t), in management confidence could whereas the commercial catch has be achieved. ranged from100 to 200t per annum. 2. Determinationof distribution of A generalised linear model of ocean large and old fishes. In NSW and beach recreational club data showed WA, large fish are found offshore that year and location were significant where spawning occurs. Since factors. A major increase in these areas are less accessible to standardised catch rate during the late anglers, the offshorezone seems to 1960's and early 1970's was observed, buffer their stocks from overfishing. followed by a decline in catch rate to The Queensland stock appears at the 1960's level. The catch rate forthe firstsight not to have this natural last decade from1982 to 1992 (the last safeguard. It was considered that year that data were available) remained a confirmationor denial of this relatively stable. possibility was a high priority. It An algorithm was developed to allocate could be determined directly by the tailor catch and effort to the ocean searching for large fish offshore, beach and other mesh fisheries. perhaps from charter vessel catches. Standardisation of catch rate through Alternatively,it would be possible generalised linear modelling did not to track onshore schools acoustic highlight any significanttrends in the ally to determine whether they data. The data, however, were very move offshoreany great distances variable. or to determine whether there is a substantial population offshore. Previous size frequencydistributions had not been weighted to sampled catch 3. Improve our understanding of size. Reanalysis of the available data the relation between fishery had little effect, but it was performance indicators and catch recommended that this method of per-unit-effort data. As with eastern weighting be used in the future. king prawn and saucer scallops in previous chapters. Again a high priority.

114 4. Sample commercial catches (both with greater efficiencythan beach seining and charter boats) in a recreational catches. representative fashion. The recommendations being developed 5. Fishery Independent Surveys; fromFRDC project (T94/161) possibly aerial surveys. This would should be noted. Commercial not resolve the offshore or ageing catches only make up a small problems and so was given a lower portion of the total catch but it is a priority than the firstfour strategies. targeted catch and can be sampled

Category Comments Commercial Catch Spatial resolution not at Management Plan scale for the ocean beach fishery. Gear type below mesh net is not specified. Effort As above. Search time not recorded. Combine crews from different endorsements either logs catch per endorsement (resulting in 2 days of effort being calculated)or as a single log (resulting in 1 day of effort being calculated). Definition of effort is unclear in multi-species and multi-endorsed fishery. This problem is exacerbated by target species and zero catches not being recorded. Catch rate As above. Recreational Catch, effort and Recreational catches larger than commercial catch rate fishery. Only once off estimates from survey, thereforeno time series available. Independent index of biomass or None. recruitment Estimates of natural mortality Estimates fromage data. Estimates of fishing mortality or biomass Estimates from age data. Input controls Limited entry, spatial and area closures, minimum legal size. Output controls None. TACC Decision rules NIA. Performance indicators None.

Table 6.6. A summary of data quality, stock assessment and management ptrogress to date for tailor.

115 6.5. REFERENCES Bade,TM 1977: Thebiology of tailor(Pomatomus saltatrixLinn.) fromthe east coast of Australia. MSc Thesis, University of Queensland, Brisbane.117 p. Barger, L.E. 1990: Age and growthof bluefishPomatomus saltatrixfrom the northern Gulf of Mexico and U.S. South Atlantic coast. Fisheries Bulletin, 88(4): 805-809. Beverton, R.J.H. & Holt, S.J. 1956. A review of methods for estimating m01tality rates in fish populations, with special references to sources of bias in catch sampling. Rapport et Proces-Verbaux des Reunions, Conseil International, pour L'Exploration scientifique de la Mer Medeterranee.. 140: 67-83. Butterworth, D.S., Punt, A.E., Borchers, D.L., Pugh, J.B. & Hughes, G.S. 1989. A Manual of mathematical techniques for linefish assessment (incorporating a report of the SANCOR Marine Linefish Program's Workshop on Population Dynamics, 4--6February 1987, Cape Town). South. AfricanNational Scientific Programmes. Report Number. 160: 81 pp. Chapman and Robson, D.S. 1960. The analysis of a catch curve. Biometry. 16(3): 354--368. Galluci, V.F., Saila, S.B., Gustafson, D.J. & Rothschild, B. 1995. Stock Assessment: Quantitative Methods and Applications for Small-Scale Fisheries. Lewis Publishers, New York. Hoenig, J.M. 1983. Emperical use of longevity data to estimate mortality rates. FisheryBulletin, 82: 898-903. Hoyle, S.D. and Sumpton, W.D. (in prep); A yield per recruit model for snapper (Pagrus auratus) using bootstrap and monte carlo methods Lenanton, R.C., Ayvazian, S.G., Pearce, A.F., Steckis, R.A., and young, G.C. 1996. Tailor (Pomatomus saltatrix) off WesternAustralia: Where does it spawn andhow are the larvae distributed? Marine and Freshwater Research, 47: 337-46. Miskiewicz, A.G., Bruce, B.D., and Dixon, P. 1996. Distribution of tailor (Pomatomus saltatrix) larvae along the coast of New South Wales, Australia. Marine and Freshwater Research, 47: 337-46. Pauly, D 1980. On the interrelationships between natural mortality, growth parameters, and mean environmental temperature in 175 fish stocks. Journal Du Conseil InternationalPour L 'Exploration De La Mer, 39(2): 175-192. Pollock, B.R. 1979. Anglers probably took the bulk of theQu eensland tailor catch. Australian Fishermen September, 1979: 31-32. Pollock, B.R. 1980. Surprisesin Queensland angling study. Australian Fisheries 39(4): 17-19. Punt, A.E. 1992. Pc-yield II user's guide. Benguela Ecology Program Report No 26. BEP, Cape Town, 36 pp. Rikhter, V.A. and Efanov,V.N. 1977. On the approaches of estimating natural mortality of fish populations.International Commission for the Northwest Atlantic Fisheries, Research Document 76/VI/8: 12pp.

116 Terceiro, M and Ross, J.L. 1993. A comparison of alternative methods for the estimation of age fromlength data for atlantic coast bluefish,Pomatomus sultatrix. Fisheries Bulletin 91: 534-549. Van der Elst, R. 1976. Game fish of the east coast of southernAfrica. I. The biology of the elf, (Pomatomus sultatrix ) (Linnaeus) in the coastal waters of of Natal. South African Association for Marine Biology Research, Oceanographic Research Institute Vol. Investigational Report No. 44.. Wilk, S.J. 1977. Biological and fisheries data on Bluefish,Pomatomus sultatrix (Linnaeus). Technical Series Report No 11. Sandy Hook Laboratory, NOAA. 56pp. Zeller, B.M., Pollock, B.R., and Williams, L.E. 1996. Aspects of the lifehistory and management of tailor (Pomatomus saltatrix) in Queensland. Marine and Freshwater Research, 47: 323-9.

117 118 common names BREAM

Acanthopagrus spp.

WHITING Sillago spp.

DUSKY FLATHEAD Platycephalusfuscus

119 7.1. PARTICIPANTS

Mr David Bateman (Recreational representative, SUNFISH, Brisbane) Dr Ian Brown (Fisheries Scientist, DPI, Deception Bay) Dr Charis Burridge (Fisheries Scientist, CSIRO, Cleveland) Mr Adam Butcher (Fisheries Scientist, DPI, Deception Bay) Mr Darren Cameron (East Coast Inshore Finfish Resource Manager, QFMA, Brisbane) Dr Tony Courtney (Fisheries Scientist, DPI, Deception Bay) Ms Cathy Dichmont (Fisheries Scientist, DPI, Deception Bay) Dr Malcolm Dunning (Fisheries Scientist, DPI, Brisbane) Dr Brendan Farthing (Mathematician, QUT, Brisbane) Dr Charles Gray (Fisheries Scientist, NSW Fisheries, Sydney) Dr Neil Gribble (FisheriesScientist, DPI, Cairns) Dr Malcolm Haddon (Head, Department of Fisheries, AMC, Launceston) Mr Ian Halliday (Fisheries Scientist, DPI, Deception Bay) Mr Jim Higgs (Programmer/Analyst of RFISH, QFMA, Brisbane) Mr Simon Hoyle (Fisheries Scientist, DPI, Deception Bay) Mr Eddie Jebreen (Fisheries Scientist, DPI, Deception Bay) Ms Kath Kelly (Fisheries Scientist, DPI, Deception Bay) Dr Don Kinsey (Subtropical FinfishMAC Chair) Dr David Mayer (Biometrician, ARI, Yeerongpilly) Mr Daryl McPhee (QCFO and Subtropical Finfish SAG Representative, Clayfield) Mr Mark McLennan (Fisheries Scientist, DPI, Deception Bay) Mr Michael O'Neill (Fisheries Scientist, DPI, Deception Bay) Ms Michelle Sellin (Fisheries Scientist, DPI, Deception Bay) Mr Jonathan Staunton-Smith (Fisheries Scientist, DPI, Deception Bay) Mr Richard Steckis (Fisheries Scientist, WA Fisheries, Perth) Mr Wayne Sumpton (Fisheries Scientist, DPI, Deception Bay) Mr Lew Williams (Fisheries Scientist, DPI, Brisbane) Ms Kate Yeomans (Fisheries Scientist, DPI, DeceptionBay)

120 7.2. INTRODUCTION Whiting, bream and flatheadhave been the basis of the inshore commercial 7.2.1. DESCRIPTIONOF fishery since the early19th century THE FISHERY (Kailola et al. 1993). The estuarine and nearshore finfish From the early 1900s there are reports fishery is the longest-established fishery of significant catches of fish, up to in Queensland, dating back to the mid 900 tonnes per year (Williams 1993). 1800s. It is a multi-species, multiple However, regular catch data by species gear fishery shared between a or species group did not become avail commercial fleetof some 300 small able until 1945, when the Queensland vessels, which land an average annual Fish Board began to record landings at catch of about 900 tonnes, and a large its regional depots along the coast. population of recreational anglers taking a catch of a similar order Recreational activity has almost of magnitude. certain} y been a feature of the exploitation of the State's inshore Indigenous communities along the coast finfishresources since the development of Queensland have been involved for of the fishingindustry. Some angling centuries in small-scale fisheries for clubs have records of yellowfin bream inshore species such as sea mullet. and whiting catches fromMoreton Bay Little informationexists on the dating back to the early 1920s. With magnitude or composition of catches the growth of population centres along fromthis sector in the southernpart of the easternseaboard, particularlyin the the State, but they were probably minor south, recreational fishing pressure has in comparison with the present-day been increasing steadily. Both commercial and recreational finfish recreational and commercial activities landings. have been made more efficientby the evolution andready availabilityof A suite of species, including yellowfin outboard motors, light-weight, trailable bream, summer whiting (two species), runabouts, off-roadfour-wheel-drive trumpeter whiting, and flathead, can be vehicles, and affordableelectronic fish considered the mainstay of this fishery. finding and navigational These species are typically sub-tropical instrumentation. and generallydo not extend the length of the Queensland coastline. This is In the commercial 'mixed' fishery (all evidenced by the fact that, over the types of fishing operation including period1988- 1994, 88% (by weight) of crabbing but excluding trawling), the the State-wide catch of these species species targeteddepend to some extent ° was derived fromlatitudes southof 22 upon thetype of fishing gear employed. 30'S (a line between Cape Clinton and However, thebulk of the mixed fishery the southerntip of the Swain Reefs). inshore/estuarinefish catch in southern Queensland is taken by mesh (gill) nets 7 .2.2. DEVELOPMENT OF THE FISHERY andtunnel nets. The main non crustacean species takenin the mixed Commercial fishingin Queensland fishery include: began in the early 1800s with • yellowfin bream settlement at the Redcliffe Peninsula, (Acanthopagrus australis); and until the tum of the century was • largely confined to the foreshoresof sand or summer whiting (Sillago ciliata); Moreton Bay and the Brisbane River.

121 • gold-lined or summer whiting The ocean beach component of the (Sillago analis); mixed fishery, which takes mainly • trumpeter or winter whiting mullet and tailor, was considered in (Sillago maculata ); a separatesection of the workshop. • dusky :flathead Also included in Figure 7 .1 are reported (Platycephalus fuscus); catches of the three important small • mullet species (Mugil cephalus, mackerel species - school mackerel M. georgii, Myxus elongatus, and (Scomberomorus queenslandicus), Liza argentea); spotted mackerel (S. munroi) and • tailor (Pomatomus saltatrix); grey or broad-barred mackerel • small mackerels (Scomberomorus (S. semifasciatus). A number of other queenslandicus, S. munroi and species, considered as by-catch on S. semifasciatus). account of sporadic occurrenceor

250.0

DNorth 200.0 •south

.c 150.0

100.0

50.0

... C: Q) Q)... u -0 ...... lo::: >, 0 ct1 >, 0 Q) ..c: ..c:. 0.0 ...... ct1 .Q u ct1 0 -0 Q) Q) E a. Q) ·;;:: Cl) en ::i a. i ..;::: � .c:: ea> a. E en .Q ..c: ·ea Q) ...... t 0 u u Cl) E C I- -0 ct1 ct1 en Q) C ::i ::i a3 rti ::i � ... ::i ::i ::i ..... en >, (.) (!:l a3 ... I ...... u.. ..J � .... ct1 ...lo::: z, C) a3 Q) ea u C) C: .... a3 ...lo::: ct1 C) C E Q) .... u ea> E CCI ...lo::: Q) CCI C :-e ...lo::: ea Q) ..c:. .c:: ....Q) u u � :.c s ea ct1 CCI I- s s � � Figure 7.1. Mean annual catch (1988-1994) of the main components of the southern estuarine fishery derived from the Queensland east coast northand south of Cape Clinton (22 °30'S). (Note Black trevally are a siganid and not a carangid).

122 relatively limited quantity rather than becoming caught in themeshes by necessarily lower economic value, are protruding finspines or gill covers, also taken. Due to time and priorities, or simply by trying to force their these species were not considered at the way through the mesh. workshop. Dusky flathead, for example, are The recreational fishery also targets usually captured in mesh nets by yellowfin bream, summer and entanglement. The existence of trumpeter whiting, and dusky flathead. vomerine teeth, pre-opercular spines, Other species caught incidentally assorted head ornamentation, and a include tailor, luderick, tarwhine, large flathead in relation to the main and dart (Pollock & Williams 1983). body trunk appears to predispose flathead to capture in nets of various Commercial quantities of whiting mesh sizes. Flathead are consequently (mainly the trumpeter or winter whiting captured in nets of a mesh size which S. maculata) are caught as bycatch in would usually not retain a more the prawn trawl fishery in Moreton Bay fusiformshaped fish with an identical and other estuarineareas. Small girth measurement. Most flathead quantities of flathead are taken as well, larger than 50 cm are captured by the but these may include species (other entanglement of several separate than P. fuscus) which are not generally meshes over each pair of pre-opercular caught in the net fishery. The trawl spines. fishery takesno appreciable by-catch of yellowfin bream. Estuarinespecies are often specifically targeted during a net shot by inshore It should be pointed out that there is mesh net fishers. Fishers are able to another fishery separately managed target particular species by considering and currently in a developmental stage factors such as mesh selectivity, bottom which specifically targets the prolific substrate, state of tide, and season. stout whiting (S. robusta) offshore in Consequently, the catch on any given depths of 25 32 m between Sandy day by a particular fisherwill tend to Cape and Bribie Island. S. robusta be dominated by one species. do not occur in the estuaries, and are therefore not part of the estuarine/ Tunnel netting is a 'draining off' inshore fishery. operation involving the use of a fixed net staked out in the intertidal zone, The commercial estuarine-inshore usually on mud-flatsin frontof mixed fisherycatch is taken mainly mangrove forestsor near the mouth of by mesh and tunnel net. Significant a river or creek. The wings of the net quantities of trumpeter whiting are are fixed in such a way as to shepherd taken as by-catch by the Moreton Bay fish towards a long sock or blind tunnel prawn trawl fleet. Some haul or seine submerged in a shallow gutter on the netting also takes place around the ebbing tide. As the tide falls, the wings foreshores of the bay for mullet, are normally dismantled so that whiting, or (with nets of smaller mesh ultimately only the tunnel remains, size in seagrass areas) for garfish. at least partly submerged in sufficient depth of water to allow the catch to Gill or mesh netting involves the swim freelyuntil they are sorted. At deployment of a light monofilament certain times of the year concentrations net in an area where fishare likely to of jellyfish (blubber) can build up be moving and may swim into the net, against the net and force it beneath the

123 surface, and drifting filamentousalgae This type of gear is used from small (blanket weed) covering the mesh can boats, the foreshore, river mouths, and reduce the net's efficiency. Tunnel man-made structures such as rock nets are not as selective as mesh nets walls, wharves, and jetties. andtend to capture a broader range Recreational anglers are not permitted of species. to use nets apartfrom bait nets (maximum length and width 16 m and Seine or haul netting is normally 3 m respectively) and cast nets conducted fromthe foreshore or beach, (maximum diameter 6 m). The generally with theaid of a small vessel maximum mesh size permitted for both to lay the net out in an arc, surrounding types of net is 28 mm to ensure that the an area of water suspected of containing catch comprises fishof a size suitable fish. The net is then hauled in to the only as bait. shore ( sometimes with the aid of a vehicle equipped with a winch) where Bream, whiting, flathead, and tailor are it is 'dried out' in very shallow water the most popular angling species in the enabling the catch to be sorted estuaries and inshore waters of southern manually. Queensland. Yellowfin bream are the main species takenby recreational At the present time it is not possible fishersin the estuarine areas of Moreton in the QFISH database to discriminate Bay, Caloundra, Jumpinpin, and between the various types of mesh Southport. These species are also and seine netting operation. This all very important components of the contributes to a significantdifficulty commercial fishcatch. in monitoring the status of these resources using catch rate (CPUE), Several attempts have been made as the differentmethods are clearly to estimate the size and species characterised by substantially composition of the recreational catch. incompatible measures of fishing effort. Pollock (1980) conducted a seriesof angler interviews at Jumpinpin and As significant quantities of trumpeter Caloundra in 1979, and estimated the whiting (S. maculata) are taken as a by recreational catch of yellowfin bream catch in the prawn trawl fishery, it is in that area to be 160 tonnes. He necessary, for completeness, to include concluded that on a regional basis otter trawls in the description of it probably exceeded the reported catching apparatus. Most trawl-caught commercial net catch of 275 tonnes trumpeter whiting are taken in the ( data from QFB Reports, averaged Moreton Bay area, where trawling is over the period1977-1980). The restricted to vessels less than 14 m commercial summer whiting catch towing (usually) twin trawls witha was estimated to be greater thanthe combined headrope length not recreational catch in southern exceeding 8 fathoms (14.6 m). Trawls Queensland. aregenerally of the Sandekan or Florida Flyer design, with minimum stretched Pollock (1980) considered that during mesh of 1.5" (38 mm). the previous decade the catch from the commercial net fisheryhad increased Therecreationa l catch fromthe only slightly, in contrast to a much estuarine/inshorefinfish fishery is taken greater rise in angling activity. Small almost exclusively by baited rod-and scale recreational creel surveys in line and handline(Kailola et al.1993), Moreton Bay in 1993 indicate thatthe with a maximum of 6 hooks per line. total recreational catch of dusky

124 flathead is at least equivalent to, Whiting are marketed locally as chilled and probably exceeds, the total or fresh whole fish or fillets. Summer commercial catch. whiting species command high prices ($6.00-6.50/kgf or mediums and There is a perception among anglers $7.50-8.00/kg for largefish [typical that decreases in their catch rates are 1996 prices for whole fish]) compared due to commercial fishing activities. to winter whiting ($2.00/kg) because Moore (1986) found that 67% of of their larger size and superior flesh Hervey Bay anglers believed that catch quality. The combined value of the rates had declined. In 57% of these summer and winter whiting catch cases the decline was attributedto too (i.e. not including trawl-caught stout manytrawlers and anglers, in 20% to whiting) is probably in excess of commercial netters, and in 19% to $1.3 million before any value-adding. trawlers. Articles in recreational fishing publications in southern Theestuarine fisherysupplies local Queensland (e.g. Steptoe 1995) southern Queensland markets with fresh clearly attribute a perceived decline flathead throughout the year. A in recreational flathead catches to significantamount of dusky flathead commercial netting activity. sourced fromthroughout Queensland is auctioned whole, freshiced, by Raptis QFMA has recently undertaken a and Sons at Colmslie, Brisbane. Prices recreational-fishery targeted telephone obtained by fishers vary between about and volunteer logbook program, but $2.50 and $7.50 per kg depending on final analyses of results areas yet demand andavailability. Based on these unavailable. There is no formal prices and the quantity of flathead mechanism for apportioning the caught, the raw value of flatheadto available catch of inshore/estuarine fishers (not including any value-added fishbetween commercial and benefits from processing and additional recreational sectors. employment) is estimatedto be between $170 000 and $500 000. The commercial sector of theestuarine fisherysupplies most of its product to 7.2.3. MANAGEMENT the local south-east Queensland market, though some is sent to Sydney, The main controls on the commercial depending on price differentials. fisheryconsist of limited licence Yellowfin bream are sold almost schemes, gear restrictions, area closures exclusively on domestic freshfish which may be total or gear-specific, and markets, usually in whole chilled form seasonal closures (Quinn 1992). The (Kailola et al 1993). Large bream Queensland fishingindustry is closed in (>25 cm) fromMoreton Bay are often the sense of 'limited entry', and most sent interstate and sold at theSydney of the individual fisheries are subject to Fish Market. On the basis of an transferable endorsements. average wholesale price (to the fisher) Commercial net fishers are subject to of $3.50 per kg, the commercial bream gear restrictions in terms of type of net, fisheryis currently worth around net length, mesh size, and drop. There $420 000. Price is size-dependent, arealso weekend closures on all rivers ranging from $3.00-3.50/kg for and creeks south of BaffleCreek, and in average sized fish to $4.50-5.00/kg Moreton Bay. for large fish.

125 Size limits apply in both commercial This rather large topic has been well and recreational sectors. Minimum described in Kerby & Brown (1994). legal sizes (total length) of the species Estimates of the growth characteristics most frequently encountered in the of yellowfin breamhave varied widely inshore fisheryare as follows: yellow among different studies (Kerby & finbream 23 cm, summer whiting Brown 1984). The work of Munro 23 cm, tarwhine 23 cm, flathead 30 cm, (1944), who used scales to determine luderick 23 cm, and 'lesser' mackerels age, suggested that growth is linear and 50 cm (Anon, 1996). There are no bag relatively slow. In contrast, Pollock limits on recreational fishers at present (1984), using length frequencyanalysis for bream, whiting, or flathead, though and tagging, estimated lengths-at-age of they are currently being considered. the firstthree age classes suggesting a Input controls on recreational fishing growth curve conformingto the von restrict gear to a prescribed number of Bertalanffymodel (Figure 7 .2). Scales fishinglines and hooks. are generally regarded nowadays as poor estimators of age, so the growth Until recently, no formal performance characteristics as estimated by Pollock indicators or reference points have yet (1984a) are presumed, until tested by been developed for any of the State's comparisonwith otolith-deriveddata, fin-fisheries.To date, only adhoe to be more realistic. analyses of commercial catch-per-unit effort have been used in an attempt to Pollock (1982) estimated von draw conclusions about the trends in the Bertalanffy growth curve parameters: stock. Such analyses have been Loo= 29.5 cm, K=0.51, and to= -0.32 hampered by lack of resolution in the years. These estimates were derived data, poorly-defined measures of from mark-recapture data. Yellowfin fishingef fort, and an inherent but bream exhibit protandrous sex inversion completely untested assumption that (Pollock 1985). catch rates provide an unbiased index of stock size. 7.2.4.b Current catch and effort levels Almost all of the commercial catch of The draft SubtropicalFinfish yellowfinbream is taken by tunnel Management Plan has identified netting or gill-netting. During 1994, appropriate performance indices and this amounted to about 96% of the reference points (whether biological, bream catch reported fromthe southern economic, or social) in order to estuarine/inshorefishery between introduce a formal mechanism for Mackay (21°00'S) and the New South assessing trends in the stock, and (most ° importantly) specifying a course of Wales border (28 30'S) (Table 7.2). management action to be takenif and The small quantity attributed to when the reference pointsare reached. trotlining is probably a reporting error, as this technique is currently only used in the spanner crab fisherywhere the 7 .2.4. BREAM bycatch is almost non-existent and very 7.2.4.a Biology unlikely to include yellowfinbream in Table 7.1 below summarises the most any case. salient points of our biological knowledge of the species.

126 Species Age at Length at Spawning Habitat Migrations Maturity Maturity (cm) Period Preferences ears Yellowfin Bream Depending on Depending on June-August Shallow turbid Annual (Acanthopagrus source: source: 17.5- estuarine spawning australis) 2, 3 m 4 f, or 20.5 cm FL, waters migration 4 18 cm FL m associated from and 21 cm FL with sea-grass estuarine for 20 cm. and feeding mangroves. grounds to surfbar entrances. Sand Whiting Depending on Depending on September- Commonly Northward ( Sil/ago ciliata) source source: February found in localised 2 or 3 or March shallow water movements 26 cm FL or ( <5 m) over a (<15km) 21-30 cm FL sandy substrate. Golden-lined 2-3 20 cm September- Shallow water Unknown Whiting (Sil/ago March (<10m). analis) Adults prefer muddy substrates. Trumpeter Whiting 1m, 2 f Depending on July-February Silty and Unknown. (Sil/ago maculata) source: or March muddy No evidence substrates. of migration 12.6 cm TL m Especially outside of and 14.1 cm common in Moreton Bay. TL for 18cm turbid areas. FL. Dusky flathead 3--4 Approx. September- Mud, silt, sand Short & long ( Platycephalus 25 cm TL March & seagrass distance fuscus) beds to migrations depths observed. of 10m

Table 7.1. Summary of biological characteristics for6 species (based on Kerby & Brown 1994) (m == male,!= female)

127 Bream - growth curves

••----• Munro (1944), scales

••----• Pollock (1982), tags, 1/f

.s::. 1 C:en G) ,_J 1

5 o;-----,------r-----....------J 0 1 2 3 4 Age (yr) Figure 7.2. Indicative growth curves for bream based on mean length-at-age data from Munro ( 1944) and Pollock ( 1982 ).

Fishing method Weight caught significantly to the reported catch of (tonnes) yellowfin bream in the southern fishery, Tunnel and gill 111.58 as the pikey bream is largely restricted netting to tropical estuaries of north Queens land and Papua New Guinea. Trotline 0.99 Not specified 3.24 Historical commercial catch data are Total 115.80 available only through the records of landings at the QFBs regional depots and Fishermen's Cooperatives. These Table 7.2. Total catch of bream by fishing extend from the Second World War method in 1994 Data source: QFISH mixed fisherytable. through to around 1980, when the Board was sold to private interests. The yellowfin bream is very much a While not accounting for the entire subtropical species, occurring in commercial catch, thesedata are of estuarine waters in the southern part of interest in thatthey do not exhibit any the State and ranging south into New consistent long-term trend (Figure 7.3). South Wales. A congeneric species There appears to have been a post-war (the pikey bream A. berda) was also decline in landings over the 15 years to probably reported as 'yellowfinbream' 1960, then a gradual increase to a priorto 1994, when it firstappeared in second peakin the late 1970s. The the QFISH database. However, long-term mean annual catch over that this would not have contributed period was 220 tonnes, but individual

128 yearly catches ranged fromabout 70 t effort occurred at Jumpinpin and in 1960 to 380 tin 1946. Caloundra surf bar spawning areas between May and August of 1960 The reported annual catch of yellowfin and 1975, and a decrease to 1980. bream during the period of the QFISH J um pinpin went from9 to 27 fish per logbook programme has also varied, angler per trip (5 year running mean), from about 220 tin 1990 to 120 tin and down to about 16 (two year mean). 1994. This variability is well within the Caloundra went from5 to 20 (5 year 'historical' range mentioned above. running mean), and declined to about 15 (two year mean). The decrease was 7.2.4.c Recreational catch trends concurrentwith a substantialincrease Records of fishingclub anglers in the in recreational fishing effort. The total whole of Moreton Bay show little quantity of bream caught did not change in themean size of bream change significantly between 1975 caught between September and April of and 1980. the years1945 to 1980 (Pollock and Williams 1983). An increase in catch Thwaites & Williams (unpublished) per unit effort occurred between 1965- also examined changes in the 66 and 1975-76, from7 fishper angler recreational catch rates of yellowfin trip to 15 (five year running mean), bream, based on records fromfishing followedby a decrease to 13 by 1980- club competitions. They found no 81. A similar increase in catch per unit significant change at theJ um pin pin or

400------, .

C1) C: C: 300 · . .c.... en ::s 200- .c.... ·-en C1) � 100- .

0 ___..,..._...._...... ULI._...... ,..,.._._...._. ..._...... _. ..._...... _._....,...., 1945 1950 1955 1960 1965 1970 1975 1980 Year

Figure 7.3. Historical records of commercial bream catch received by the Queensland Fish Board at all branches and depots. Data source: Queensland Fish Board Annual Reports.

129 Southport surf bars, or in the estuarine Commercial effort levels are not yet areas of Moreton Bay, from the mid available by gear type for the entire seventies to 1991. However, they noted State. Neither catch nor effortlevels a decline at the Caloundra surf bar are available for the recreational sector. between a peak of 29 bream per angler Sillago ciliata is found along the east day (three year running mean) in 1974 coast of Australia from Cape York, to 10-12 fishper angler day in the mid Queensland, south to eastern Victoria to late eighties. and northernTasmania. S. analis is mainly restrictedto north Australian 7 .2.5. WHITJNG waters, as far south as Shark Bay in 7.2.5.a Development of the fishery Western Australia and Moreton Bay in Queensland. Almost all of the commercial mixed fisherycatch of whiting is taken by The mixed fisheryfor whiting extends beach seining, tunnel netting, or gill fromnorth of Mackay to the New South netting. During 1994, this amounted to Wales border, but almost all (95% by about 96% of the whiting catch reported weight) of the State's reported catch from the southern estuarine/inshore comes from between Baffle Creek and fisherybetween Mackay (21 °00' S) and the New South Wales border. The the New South Wales border (28°30' S) majority of the catch is evenly divided (Table 7.3 ). The small quantity between Hervey Bay (27%) and attributed to trotlining is probably a northernMoreton Bay, including reporting error, as this technique is Pumicestone Passage up to October currently only used in the spanner crab 1995, (30%). Significant proportions fishery where the bycatch is almost of the 1994 catch also came from the non-existent and very unlikely to Bundaberg north Hervey Bay region include whiting in any case. (9%), Tin Can Bay (16%), and south Moreton Bay (6.5%). Fishing method Catch (tonnes) Gill/tunnel netting 166.68 Historical commercial catch data are Trotlining 0.02 available only through the records of Unspecified 8.27 landings at QFBs regional depots and method Fishermen's Cooperatives. These Trawl 2 507.77 extend fromthe Second World War (inc. 5. robusta) through to around 1980, when the TOTAL 2 682.74 Board was sold to private interests. Table 7.3. Total commercial net and line While not accounting for the entire (mixed fishery)and trawl fisherycatch of commercial catch, these data are of whitingin 1994 between Mackay (21 °S) and interest in that they do not exhibit any the New SouthWales border (28°5'S). Data consistent long-term trend (Figure 7.4). sources: QFISH mixed fishery and trawl fisherytables.

130 The reported mixed fishery annual 7.2.5.b Recreational catch trends catch of whiting during the period of Thwaites & Williams (1994) examined the QFISH logbook programme has summer whiting competition catch also varied, from about 320 tin 1988 to records of anglers from 12 recreational 175 tin 1994 (Figure 7.5).Allowing for fishingclubs at fivepopular fishing the fact that the historicaldata includes locations in south-east Queensland prawn bycatch of S. maculata, this (Inskip Point, Bribie and Moreton variability is probably within the Islands, Jumpinpin, and Southport). All 'historical' range mentioned above. sites except Inskip Point experienced an However, the observed trend is a increase in the average number of decline, by just over 45% since 1988. summer whiting caught per anglerday Catch per unit effort, crudely estimated between 1959 and the 1970s. This was using the 'boat day' as the unit of followedby a decline between 1975 fishing effort, has remained roughly and 1991 at Bribie Island by about constant during this period, since effort 10 fishper angler day. There was no has also decreased. significantchange at theother four sites.

500 .

C C 400- 0 . :e. .c en aoo- as:::, . (.) .c ·-en . � 100- .

1945 1950 1955 1960 1965 1970 1975 1980 Vear

Figure7.4 Historical records of commercial whiting catch r�ceived by the Q�eensland Fish Board at all branches and depots. All whiting species and fishing methods are included. Data source: Queensland Fish Board Annual Reports.

131 ...... tn 35 30 �ea Cl) A, "C C / C 30 / ' .... 0 / ' 25 ea .... � ' 0 25 ' ' � .s:::; -20 Cl) C, u :::, 20 Catch ea 0--0 C, CPUE 0--0 15 .s:::;.... 15 C, 10 :::, ·-Cl) c. 3: 10 0 50 5

0 0 1988 1989 1990 1991 1992 1993 1994 Vear

Figure 7.5 Annual trends in commercial net catch and cpue (ie. excluding trawl catch)for all whiting species between Mackay and the New South Wales border. Data source: QFISH summary table.

7.2.5.c Biology commonly found in northernNew The basic biological informationf or South Wales and southern Queensland whiting are described in Table 7 .1. (McKay 1985). S. analis is mainly restricted to north Australian waters. It Thirty-one species of the family occurs in Shark Bay, WesternAustralia, Sillaginidae are recognised (McKay along the NT and Eastern Queensland 1992). The name 'summer whiting' coastlines, and southwards along the includes the species Sillago ciliata and east coast of Queensland to Moreton Sillago analis. S. ciliata is known also Bay. It is also found on the southern as sand whiting and bluenose whiting. coast of New Guinea (McKay 1985). S. analis is known also as the golden S. maculata is foundoff the east coast lined whiting, Tin Can Bay whiting, of Australia fromLizard Island, north and rough scaled whiting. Queensland, to Narooma in southern New South Wales, and on the coast of The trumpeter whiting (Sillago WesternAustralia (Kailola et al 1993). maculata) is also known as the winter It is also common on the coasts of east whiting, diver whiting, and spotted Africa, the Philippines, the East Indies, whiting. There are three subspecies of and China (McKay 1985). S. robusta S. maculata, of which only one, S. m. occurs throughout inshore marine maculata, occurs on the east coast. waters offsouth-eastern Australia. Sillaginids are widely distributed in the Estimates of the growth characteristics western Pacific andthe Indian Ocean. of S. ciliata have been produced for S. ciliata is distributed fromPapua both New South Wales (Cleland 1947) New Guinea to Tasmania, but is most

132 and Queensland (Dredge 1976) (male) fromthe Botany Bay (New populations, and they appear to be very South Wales) population. similar (Figure 7.6). However, Cleland did not differentiate between S. ciliata S. analis grows to a maximum length and S. analis, which must reduce of 45 cm TL (McKay 1992). Length confidencein his estimates. frequencydistributions, however, are difficult to interpret, probably as a Cleland's estimates are based on scales, result of the species' extended spawn while Dredge used otoliths. Dredge ing period (Gunn 1978, Kerby and found both scales and otoliths difficult Brown 1994). to interpret, and concluded that there are two distinct phases of growth in Maclean (1968) used growth checks S. ciliata. He interpreted fish smaller on scales to determine length at age for than 20 cm separatelyfrom larger fish. S. maculata, while Weng (1994) used otoliths, scales, and length frequency The maximum recorded length for analysis (Figure 7.7). S. maculata can S. ciliata is 51 cm (McKay 1992), but grow to about 30 cm totallength this is exceptional. Young (1989) at the (Dixon et al 1987), although Weng Southport broadwater recorded maxima ( 1994) found maxima of 22 cm FL for of 38 cm (female) and 36 cm (male), males, and 27 cm FL for females. while Burchmore et al (1988) reported Maclean ( 1968) reported a maximum maxima of 31 cm (female)and 40 cm size of 22 cm FLfor both sexes.

Sand whiting - growth curves 40

35 •i----•• Cleland (1947), scales • • Dredge (1976), otoliths 30

..I 15

0 0 2 3 4 Age (yr) Figure7 .6 Indicative growth curves for S. ciliata (sand whiting) from Cleland ( 1947) and Dredge ( 1976). Curves are based only on reported mean lengths-at-age, for a limited range of ages. Lengthsare fork length.

133 7 .2.6. FLATHEAD The dusky flathead(Platycephalus fuscus) occurs in estuarine waters from 7.2.6.a Exploitation History Cairns in Queensland to the Gippsland Almost all of the commercial catch of Lakesin eastern Victoria (Kailola et al flatheadis takenby mesh and tunnel 1993). netting. During 1994, this amounted to about 94% of the flatheadcatch The fisheryfor flathP,::icl P,XtP.nds from reported fromthe southern around Townsville to the New South estuarine/inshore fishery between Wales border, but almost all (98% by Mackay (21 °00'S) and the New South weight) of the State's reported catch Wales border (28°30' S) (Table 7.4). comes from between Cape Clinton and The small quantity attributed to the New South Wales border, and the trotlining is probably a reporting error, bulk of the catch (53% in 1994) comes as this technique is currently only used fromnorthern Moreton Bay, including in the spanner crab fishery where the Pumicestone Passage. A significant bycatch is almost non-existent and very proportion of the 1994 catch (21 %) unlikely to include flathead in any case. came fromthe Tin Can Bay area. Smaller catches were taken in Hervey Commercial effort levels are not yet Bay (24.5°S to 25.5°S), in the available by gear type for the entire Caloundra area, and in southern state. Neither catch nor effort levels are Moreton Bay. available for the recreational sector.

••---• Maclean (1968), scales 25 e----•Weng (1993), scales+ 1/f

0 0 1 2 3 4 Age (yr)

Figure 7.7. Indicative growth curves for S. maculata ( trumpeter whiting) fromMaclean (1968) (total length) and Weng (1994) (fork length). The Maclean curve is fitted to reported mean lengths-at-age, while in the latter case the reported von Bertalanffy growth parameters were used to plot the curve.

134 Method Catch (tonnes) through to around 1980, and do not Line fishing 0.05 account for the entire commercial catch. Tunnel and gill netting 42.43 Flathead data also do not exhibit any Trotline 0.18 consistent long-term trend (Figure 7 .8). Unspecified method 2.32 Following a period of stable catches Grand Total 44.97 between 1946 and1959 averaging 91 t, catches up to 1980 averaged 71 t (apart from peaks in 1974 and1975 of about Table 7.4 Total catch of flatheadby fishing 100 t). The long-term mean annual method in 1994 between Mackay (21.0°S) and catch over that period was 77 t, but the New South Wales border (28.5°S). Data individual yearly catches ranged from source: QFISH mixed fishery database. about 53 tin 1970 to 106 tin 1958.

As for bream and whiting, historical The reported annual catch of flathead commercial catch data are available during the periodof the QFISH logbook only fromthe Second World War programme has also varied, fromabout 83 tin 1989 to 45 tin 1994, with a declining trend (Figure 7.9).

- 125 C ....C 100- .c 75- a, ea::::, - 0 - .c.... 50 a, .. "ai:s: 25--

0 1945 1950 1955 1960 1965 1970 1975 1980 Year

Figure 7.8 Historical records of commercial jlathead catch received by the Queensland Fish Board at all branches and depots. All species ofjlathead are included. Data source: Queensland Fish Board Annual Reports.

135 .-. .-. u, 100 10 >- a, as C "C C ..... 0 80 8 as ..... 0 .._.. fa--- .c ..... / .r:. / -o.___ i. /- �-0- - a, C)::s 60 a 6 c. as C) .....(.) .._..� .c 40 4 w ·-C) ::::> a, Catch 0---0 D.. CJ � 20 CPUE 0--0 2

0 0 1988 1989 1990 1991 1992 1993 1994 Vear

Figure 7.9 Annual trends in the mixed fisherycatch and cpue offlat head, for all areas between Mackay and the New South Wales border. Data source: QFISH summarytables.

7.2.6.b Biology and behaviour accompanying graph. These estimates Table 7.1 describes the basic biology vary somewhat fromthose of Dredge of flathead. (1976), possibly because of the difference in interpretation of the 7.2. 6.c Growth first annulus, which is often difficult to detect. Dredge (1976) estimated the growth of dusky flatheadusing otolith ageing techniques and length frequency 7.3. BREAM, WHITING analyses. Scales were not used for AND FLATHEAD TASK aging because growth rings are unclear. The mean lengths of fish with 2, 3, 4, GROUPS and 5 otolith annuli were 23, 33, 44, 1. Investigationof catch rates from and 52 cm respectively. Dredge (1976) CFISH specially in terms of fishing foundno significant difference in method and location length-at-age between male and female dusky flathead. 2. Estimate relative commercial to recreational catch Preliminary estimates of the von 3. Discuss andanalyse theusefulness Bertalanffy growth parameters from of the recreational tagging database recent studies at theSouthern Fisheries with special emphasis on Centre are as follows: L"" = 84.4 cm, movement, growth anddistribution k = .205 and to= -0.94 (Figure 7.10; of effort Cameron unpubl. data). Computed lengths-at-age based on these parameter 4. Re-examine growth rate estimates estimates are presented in the

136 Dusk flathead - rowth curve 90 80 • • Cameron unpubl. data; otoliths .-. 70 E 60 .c 50 C: 40 30 20 10 0 0 1 2 3 4 5 6 7 8 9 10 11 12 Age (yr) Figure 7.10 Growth curvefor Platycephalusfuscusbased on computed lengths-at-age (D. Cameron, unpubl. data). 7.3.1. INVESTIGATION OF CATCH of bream, they assumed that most of RATES FROM CFISH the data were mainly catches of This group investigated commercial yellowfin bream. catch rates in terms of effort The latitudinal range of data processed standardisation. A further objective was 24.50 degrees decimal to the New was to determine how often the South Wales border. The log of daily proposed management plan trigger catch rate was modelled in a General points would be activated. Linear Model (GENSTAT) using a Analysis of biological status of stocks surrogate of vessel type known as rely on a knowledge of the relationship 'vessel sequence number', year, month, between effort and catch. However, as half degree square QFISH grids and net in many fisheries, this relationship is type (whether >800 and <=800m net affected by factors other than biomass length). There were therefore about changes. Effort standardisation is an 11 000 records with net lengths � 800m objective method of investigating the and about 29 000 records > 800m various factors such as vessel A plot of standardised and characteristics, gear used etc. that unstandardised catch rates for bream affecteffort. are given in Figure 7.11. This model This group only had time to complete explains 37% of the variance. The an analysis of bream. Although the standardised trend is much less database does not always distinguish optimistic than that of the properly between the differentspecies unstandardised trends. The months

137 April to August are significant standardisation of bream and flathead (t, p<0.001) and vessel sequence was completed by Ms Dichmont and number explains most of the variance. will be reported on in detail in Brown Given these standardised catch rates et al. (in prep). for bream, the proposed QFMA management rules would have been 7.3.2. ESTIMATERELATIVE triggered twice between 1988 ,md present. COMMERCIAL AND RECREATIONAL CATCH This group recommended that this The species resolution of the method of removing factors that affect recreational and commercial databases catch rates, other than biomass, is was investigated. This is particularly a essential to correctinterpretation of problem in species such as whiting as biomass trends. In the analysis of they may be difficultto tell apart, but bream, for example, the standardised have very distinct life histories. It may index resulted in a differenttrendline, therefore be questionable whether they which would therefore be open to a could be managed as a single unit. This potentially different interpretation and group was therefore also asked to management advice. highlight the fishcategories that they Effort standardisation would be were unable to break intospecies and extremely difficult on whiting, as the that may be very differentin their data consists of several species with habitat and biological. characteristics. differentlife histories. However, the They also investigated the relative catch group did recommend that this analysis participation of the recreational and be extended to flathead. Effort commercial sectors.

1.6 1.4 w :::, 1.2 D.. 0 1 ·-Q) 0.8 ea 0.6 - - -0- - Stand Q) cc: 0.4 • Unstand 800m 0.2 z!!s 0 co 0) T"" C\I ('t') -.:::I" LOCD co co 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) 0) T"" T"" T"" T"" T"" T"" T"" T"" T"" T"" Year

Figure 7.11. Standardised catch rates ( 'Stand') and unstandardised catch rates of yellow.finbream for net length ::;800m and > 800 m ( 'Unstand <800m' and 'Unstand >800m'). Note that the indices have been scaled to unity in 1988.

138 7.3.2.a Recreational catch to divide the recreational catch into QFMA recently initiated a telephone species and only broad groupings can and logbook survey of the recreational be used (e.g. Table 7.5). An example of catch of major fish species within the method of calculation is given in Queensland. This survey is a first Table 7.6. statewide attempt at investigating Quarter Daily Correction Corrected recreational catch and participation Harvest factor harvest demographics. The preliminaryresults 1st 2135 1 2135 were kindly made available to the 2nd 4099 0.9 4554 workshop. 3rd 2614 0.75 3485 4th 1038 0.5 2076 A first estimate of total recreational catches for each broad fishcategory Table 7 .6. Bream as an example of relevant to this workshop session, are recreationalcatch calculation. Correction factor represented in Table 7.5. Average fish adjust for the decline in participation rate. weights were taken fromthe length Total harvest is therefore estimated frequencydata of the recreational creel from the diary data as the sum of surveys completed in 3 estuaries by the corrected quarterly harvests. The total Coastal Streams Project (Queensland recreational harvest becomes scaled by Government New Initiative, Fish the ratio of diaryparticipants to total Management & Protection, recreational fishers. 'Assessment of Fish Stocks in Coastal Streams', SouthernFisheries Centre, Of the three groups, Bream spp. Deception Bay, 1997-1999). Since contributes the most to the recreational whiting catch consists of both summer catch by mass. Flathead and whiting and winter whiting, average individual were equally important. This work is weights were calculated by weighting extremely preliminaryand has yet to individual summer and winter whiting be analysed in detail by the Bureau of into weights by the proportion of Statistics for QFMA and released in identifiedcatch in the RFISH database. a detailed report. The calculation took into consideration that there was a decline in participation 7.3.2.b Resolution of species in the recreational data collection Many fishersare unable to differentiate program over time. It was not possible between the differentspecies of bream, Average Average Max. whiting or flathead. As a result, the individual total total resolution of the commercial and weight (kg) catch catch recreational databases is not at a (t) (t) species level. The objective of this Bream 0.278 455 681 group is to analyse whether it is Flathead 0.63 292 437 possible to use the database to evaluate Whiting 0.049 273 410 the catches to the species level or to (spp.) produce an algorithm forspecies Table 7 .5. Estimated average individual separation. If this separation is not weight for recreational catch frompreliminary possible, to discuss the importance of recreational phone and logbook survey by this informationespecially with respect QFMA. Whiting (spp.) includes the weighted to management needs. average of summer and winter whiting. The catch by species category for bream, whiting and flathead are given in Table 7.7. It is clearthat several

139 species categories only have data for whiting. The catches of the mesh net some of the years as thespecies list was licensed fisherywere interpreted to extended over time. In terms of bream, comprise mostly summer whiting. 99.4% of the total catch data have been recorded in the 'bream unspecified' Only one category of flathead is category. It was the general opinion available foruse in the CFISH of the group that most of this catch database; 'flathead- unspecified' and consists of yellowfin bream, but, there it was agreed that this would consist was no obvious method of dividing the mostly of dusky flathead catch. A catch without error. total catch from 1988 to present is about 650t. Most of the whiting catch was recorded in the 'whiting - unspecified' and 7.3.2.c Fishing sector separation 'whiting - summer' categories. The The commercial bream catch in group stated that most of the summer Queensland waters is about 175t whiting catch would have been (and decreasing). Annual catches from recorded in the 'sand', 'summer' and the recreational sector were estimate 'unspecified' categories. Excluded from at 455t (maximum 700t) for the Table 7.7 is a sizeable trawl by-catch of recreational sector. The recreational whiting ( other than stout whiting) of to commercial catch ratio for bream is about 60 t. Most scientists in this group thereforein the order of 3 to 1. were in agreement that more than 90% of this trawl catch would be winter

1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 TOTAL Bream - Tarwhine <1 1 1 2 Bream 214 222 223 174 176 146 131 214 149 171 1820 Unspecified Bream - Yellowfin <1 <1 <1 9 9 Flathead 68 86 78 71 64 60 54 52 53 62 648 Unspecified Whiting - 37 6 2 2 1 <1 <1 <1 5 4 58 Trumpeter/Diver Whiting - Northern <1 <1 1 1 Whiting - Sand 2 <1 3 6 3 15 Whiting - Summer 290 305 308 269 186 19 18 21 27 33 1476 Whiting <1 <1 10 106 249 195 182 211 279 1233 Unspecified

Table 7.7. Total yearly catch by CFISH species category for bream, whiting and flathead. Note that missing data means that this category was not available at the time or that it was not used.

140 2

0 "' "' a, 0 "' 0) "' C "' a, 0 "' "' a, 0 "' 0 "' N N N ..."' "' :;: \11 "'tll i "' "' "' iii � ;:: ill ill gi ... 0) !fl ii; gi ;!! ;!! ;!! � ;!! � ;!! ;!! ;!! ;!! � � � � ;!! � � � ;!! � � � � � � I � §l Year

Figure 7.12. Catch per unit effort (no. fish per angler per trip) of whiting from recreational club data caught at Jumpinpin.

Dusky flatheadcommercial catches were able to identify which whiting range from75 to 40t, whereas the species theycaught. Extrapolating this recreational catch is estimated as an ratio (8% summer and 92% winter average of 292t (maximum estimate whiting), the recreational summer 437t). The resultant recreational to whiting catch in Moreton Bay would be commercial catch ratiois 9 to 1. about 90t and that of the winter whiting about 180t. However, the DPI New Some participants in the RFISH Initiative 'Assessment of Fish Stocks in recreational survey in Moreton Bay Inshore and Coastal Streams' Project

25 �.... G)... g' 20 ....<( .c � 15

0 "' "' a, C "' .... "' (0 a, 0 N (0 "' 0 "' "' a, 0 "' (0 a, C N a, 0 N N OJ "' "' ... "' "' .,. "' "' "' "' iii "' (0 ...... ;!: ...... a, "' iii ill "' "' :" � :" :" :" �gi :" :" "' ! ! :" :" � ill :!l :" :" :" :" :" :" :" :" :" :" :" :" "' :"� :" Year

Figure 7.13. Bream catch per unit effort (no. fish per angler per fishingtrip) for Jumpinpin recreational data.

141 estimates the catches in the 3 other recreational tagging programwith the major estuaries as being lO0t of major financial contributor being DPI. summer whiting and 200t of winter Sportfishers involved in this tagging whiting. Generally, the recreational program target several species such as catch for winter whiting far outweighs flathead,whiting, bream, mackerel, the commercial line, net and trawl mangrove jack etc. The tag-return data catch, whereas the summer whiting were kindly made available to t.11c fishingsectors seem equal in terms of workshop. catch. However, the above rules of thumb are very subjective and it is The ANSA bream tagging database was really impossible to use RFISH or investigated to determine whether it CFISH to discriminate between the would be of use for: catch of differentwhiting species. • growth estimation; Unfortunately, these species have very different life histories and grow to • movement analysis; differentmaximum sizes. • estimation of mortality rates. Recreational club data were made The database contains about 6 040 availablepooled over fishing clubs. bream records, of which 4 984 had both The relative importance of whiting and total and caudal fork length recorded. bream has changed much in the last few Of the released individuals, 305 were decades. Bream certainly has always recaptured (236 recorded as total length been a more prized fishin the catch. and 50 as fork length). Most of the tags had been released in June. A plot of bream and whiting catch rate from club data (catch per angler per 7.3.3.a Growth estimation trip) in the Jumpinpin area indicates A total length (mm) versus total weight large changes fromone year to the next (g) conversion were calculated 2 (Figure 7.12 and 7.13). No overall (r =0.578) with the result: trend is obvious, although an increase in bream catch rates from the 1950s to W(g) 0.000485L:·339 the 1970s can be discerned (Fig. 7.13) and total length versus fork length The group was concerned that they 2 were unable to divide the catch by conversions (r =0.9726) are: species which have distinct life L histories. No scientific algorithm could 1 =0.9118L1 +0.5254 be developed, only approximate rules of thumb. This applies particularly to Weights of animals were estimated to whiting spp. the nearest10 grams. 7.3.3. DISCUSS AND ANALYSE THE In termsof changes in length between USEFULNESS OF THE tag and release periods, much of the RECREATIONAL TAGGING data showed negative growth (Figure DATABASE 7.14). Furthermore, it shows that most of the animals were returnedwithin a The Australian National Sportfishing month of their release. Association (ANSA) Queensland has been involved in a Statewide

142 120 -E 100 • E 80 • .c 60 • C, 0 • • C: 4 •• • • � 20 • •• C: •• • 0 • G) C, -20 2 3 4 C: -40 0 -60 -80 Months at large

Figure 7.14. Growth increment of returnedbream in relation to time between tag and recapture. Most of tagged animals are returned within a month of release.

In an attempt to reduce the records with removed and classified as errors. An negative growth, the group focused on exponential decline model was fitted recapture records based on data in to the number of returns over timeat which the same angler had done both large. Since most animals were the release and recapture measurements. recaptured after a very short period, this However, no substantial change to analysis does not have much credibility. Figure 7.14 was observed. 7.3.3.d Conclusions Length increments :::;; -10 mm were This group concluded that the ANSA removed and classified as errors. A tagging data would be excellent for standard least squares von Bertalanffy movement studies, studies of the growth model was fittedto the data, but distribution of effort and a comparison very large residuals were observed of length frequencieswith other data (Figure 7.15). It was concluded by this sets (although this data does not have group that the tagging data would not good resolution). However, the dataset be very useful for growth analysis. is not useful for growth analysis and estimates of mortality will also produce 7.3.3.b Movement spurious results. Most of the animals remained within a short distance of their release sites, with 7.3.4. RE-EXAMINE GROWTH only a few individuals moving more RATE ESTIMATES than 30 km (Figure 7 .16). This group Various sources of otolith ageing data is recommended that the movement data available, the most important of which would be extremely useful for more in is that fromthe 'Integrated Stock depth movement studies, although most Assessment and Monitoring Project' of the recaptures were taken less than a (ISAMP FRDCT94/161). This group month after release. only had time to investigate flathead data. Flathead was chosen as the ageing 7.3.3.c Mortality estimation of this animal is known to have low In order to calculate total mortality, variation and relatively good length increments :::;; -10 mm were readability. Two issues were discussed:

143 100 80 • • 60 • • U) 40

::, "C 20 • U) 0 Cl) a: -201 0 4 0 -40 ·60 • -80 Initial Length

Figure 7.15. Residuals of a von Bertalanffygrowth model fitted to bream tag-return data. All records that showed a negative growth of 1 cm or more had been removed.

70 60 - 50 40 Cl) :::s 30 C"' Cl) 20 LL 10 0 0 0 0 0 0 0 0 0 0 0 co CJ) C\I LO CX) T""" """ I"-- T""" T""" T""" C\I C\I C\I Distance (km)

Figure 7.16. Frequency plot of distance (km) moved between tag and recapture by bream.

• otolith readability; discuss several ways of addressing any discrepancies between the readers. • analysis of growth curves and the One method is to use all the readings effect of gear selectivity on the age including those that have ageing structure. disagreement between readers. Another method is to weight fish that have been 7.3.4.a Otolith readability aged more than once so that each A birth date was given to each fish. In otolith would receive equal weighting. the project, two readers aged each The traditional way of achieving this otolith. The group therefore had to goal is to use an 'agreed' age between

144 two the readers. On the occasion where The group also discussed the effectof the two readers disagree, a third reader gear selectivity on the flathead age data. is used. Thereafter, one rejects any This aspect is important, as the data otoliths in which no agreement was represent the commercial net catch and possible. Another practice is to use a not the population as a whole. Smaller readability index. However, the group (and maybe larger) animals may be highlighted that a potential problem unselected. In the NSW commercial could exists if one is consistently net fishery, there are fewer older fish rejecting fast growers or certain age than in the Queensland fishery. Age classes because they are more structures between gear types differed difficultto age. in Queensland. Tunnel netters caught larger fish, and therefore older age 7.3.4.b Ageing analysis classes were represented in their catch. The following data was combined to Recreational fishers tendto target larger produce a growth curve for each sex: fish,and there are also clear signs of Unsexed O and 1-year olds (from the size selectivity in the mesh net fishery. Project 'Monitoring the Effectsof Restocking the Maroochy Estuary') Since not all 3-year-old animals are augmented for both males and females fully selected by the fishery, their data. This method assumes that there inclusion in the dataset could bias the are no differences in the growth rate of resultant growth curves. Similar unsexed (small) animals. Data for each problems exist in the fishery curve was therefore from independent data. Because the survey method selected fish less that 30 cm, • unsexed growth data; the mean size of the 2 year olds was • 2 and 3-year old fish from the biased downwards as some 2 year olds ISAMP dependent survey data. clearly exceed 30 cm. These animals were sexed and A birth date for each fish was decided therefore were added to the on the basis of marginal increment and relevant curve. date of spawning. Both methods

90 • • • • -E 80 E 70 60 - 50 � 40 m 30 - 20 10 0 0 2 4 6 8 10 12 Age (years)

Figure 7.17. Length-at-age curve for femaleflathead. Von Bertalanffygrowth function modelled with unweighted (solid line) data.

145 70 -E 60 E - 50 40 C: � 30 'iii 20 I- 10 0 0 2 4 6 8 10 Age (years)

Figure7.18. Von Bertalanffy Growth function modelled with unequal weighting (solid line) from commercial flatheadma le length and ototlith data.

resulted in the same birthdate would also improve the fit. There was (September - October). The group some discussion as to the relevance of chose 1 October as the nominal using a Von Bertalanffygrowth curve birthday,and converted each annual and the general conclusion is that a age to age in days. Schnute growth curve should also be investigated. Both methods of analysis discussed above were used. The firstwas The results demonstrate that flathead utilising all otoliths readings (including show a clear sexual difference in those readings fromotoliths that have growth rates (Table 7.7). Females been aged more than once i.e. reach a larger L"'°compared to males 'unweighted'). The other method only and males are slower growing. included readings from otoliths that had been read once or twice. For those that Parameter Weiohtino Males Females l,.(cm) Eaual 78.2 80.9 had been read more than twice, two Unequal 81.17 82.0 readings were selected randomly and K (vea(1) Equal 0.158 0.247 included in the analysis ('weighted'). Unequal 0.145 0.237 This process improved the fit to (year) Equal -1.17 -0.396 considerably for both sexes Uneaual -1.303 -0.418 (Figure 7.17 and 7.18). Table 7.7. Von Bertalanffygrowth parameter The sex based von Bertalanffy fits estimates giving otoliths equal or unequal were very sensitive to to mainly because weighting. there was poor representation of smaller size classes. As a result, the group recommended that smaller fishshould be aged to improve the fit. For the same reason, more larger/older fish

146 Total natural mortality (Z) was proposed QFMA draft Management estimated fromcatch curves Plan trigger points would have been (Ln(frequency)versus age) as 1.5 and activated twice since 1988. 1 0.97 year- respectively. The group The species resolution of the discussed the possible reasons forthis recreational and commercial database difference in total mortality between the was investigated. This is an important sexes. A possible explanation is that, as problem to address in species such as femalesgrow fasterand larger fishhave summer (of which there are two species lower selectivity than smaller fish, Sillago ciliata and S. analis) and winter femalesmay escape the 'window of whiting as they are difficultto tell apart, predation' faster than males. A model but have very differentlife histories. was used to simulate this hypothesis Yellowfin bream and dusky flathead which explained the catch curve very catches tended to fall mostly within the well but not the totalobserved catch. "bream-unspecified"and "flathead An alternativehypothesis is that unspecified" categories, however, behaviour differs between the sexes. algorithms accepted by the workshop This is based on an observation that in were utilised to define the catch of New South Wales waters a greater these species. Unfortunately, no proportion of male flatheadthan female satisfactory method of separating the migrate offshore. commercialfrom the recreational catch was devised forwhiting. The followingwas recommended: • Make contact with a scientist within Of the major groups of species under the University of Queensland to investigation in this section that are determine the seasonal presence of caught by the recreational fishery, larvaeof each species. This may bream spp. contributes most catch by help in determining an appropriate weight. A preliminaryestimate of total birthdate . recreational catch fromthe state-wide QFMArecreational survey undertaken • Flathead can be aged successfully in 1998 was calculated. and otolith ages should continue to be used in a monitoring program. The AustralianNational Sportfishing Association recreational tagging • The bream ageing data are database was investigated and analysed reasonable, whereas the whiting to determine whether it would be of use ageing was difficult, very variable forgrowth estimation, movement and error prone. analysis and estimation of mortality • The model hypothesised regarding rates. A large number of animals femaleflathead and the window of showed substantial shrinkage and this predation, is plausible andshould be would be most likely due to developed further. measurement error. It was concluded that thedatabase would not be useful 7.4. CONCLUSIONS for growth rate estimation. Also, since most of the animals were recaptured 7 .4.1. GROUP REPORTS after only a very short time period, analysis of mortality rates would not Standardisedtrends of commercial have much credibility. The data was bream catch rate gives a much less seen as being potentially extremely optimistic view of the fishery than the usefulfor movement studies. unstandardiseddata. Given the declining standardised trend, the

147 Length-at-age curves were analysed for effort in a multi-endorsed andmulti flathead by combining all datasets. A species fishery, fewrecreational catch birth date foreach fish was decided on or catch rate time series, and neither the basis of marginal increment and target species nor zero catches being date of spawning. Two methods of recorded, little can be said about the analysingthe data were presented. sustainabilityof catches. Flathead sho\v a clear sexual dimorphism in growth rates with 7.4.2. PROGRESS TO DATE females reaching larger average A summaryof progress to date with maximum lengths. Small and very regard to data quality, stock assessment large fish should be targeted in the and management knowledge is given future to improve the fit. in Table 7.8. Due to the lack of resolution of the data into species, the difficultywith defining

Category Comments Commercial Catch Lack of resolution to species for yellowfin bream and especially, summer and winter whiting. Gear type below mesh net is not specified. Effort As above. Search time not recorded. Definition of effort is unclear in multi- species and multi-endorsed fishery. This problem is exacerbated by target species and zero catches not being recorded. Catch rate As above. Recreational Catch, effort Generally recreational catches larger or and catch rate the same size as the commercial fishery. Only once off estimates from survey, thereforeno time series available. Independent index of biomass or None. recruitment Estimates of natural mortality Estimates fromage data. Estimates of fishing mortality or Estimates fromage data. biomass Inputcontrols Limited entry, spatial and area closures, minimum legal size. Output controls None. TACC Decision rules NIA. Performance indicators None.

Table 7.8. A summary of progress to date with respect to data quality, stock assessment and management knowledge forbream, whiting and tailor.

148 7 .4.3. MONITORING, RESEARCH estimates of total mortality and DIRECTION AND PRIORITIES stock status for these species. Table 7 .9 gives a review of the data b. RFISHdata should be validated sources analysed in this workshop with creel surveys (on-water) to get session. Broad conclusions are catch at age and length-frequencies. given below. c. Improve our understanding of a. The ageing and length frequency the relation between fishery collections of bream and flathead performanceindicato rs and catch should continue, as this provides per-unit-effort data, as with eastern insight into the status of these king prawn, saucer scallops,and stocks. The ageing of whiting was tailor above. However, there is a much less successful and would not need to breakthe catch down to prove as valuable. In the long term, species level. This should be a ageing information could provide high priority.

149 Data Source Data Type Comments Bream Flathead Whiting CFISH Catch In future, need to Yes Yes No define gear type Need to assume -No obvious used, rather than most data in mechanism of use algorithm. combined 'bream differentiating Need to use -tarwine', between different st1mdardisedcatch 'unspecified' and whiting species, rates. 'yellowfin' which have very categories are differentlife yellowfin bream. histories. -Change logbook to identify between species? -Need to include trawlbv-cateh. Effort Difficultto define effortin the multi- (j a (j species net fishery. RFISH Telephone Provides an Yes Yes Yes & estimate of Can't differentiate Can't differentiate Logbook recreational catch, between species, between species, survey which is extremely but still extremely but still extremely useful. Cost useful. useful. recovery in the formof a recreational licence was suggested as the survey is expensive. Should ground truth this data with on-water creel surveys. Recreational Catch rate Very useful for Yes No Yes club observing catch rate Most of the data Not very important trends over a long forbream. component of the time period (1992 club catch. Mainly onwards). summer whiting. ANSA Tagging Yes: Movement Yes Yes Yes studies, Yes: Distribution of effort Yes: Comparison of broad-scalelength- frequencies No: Growth analysis No: Mortality estimates. Various Otolith Useful forageing Medium Yes No FRDC,PPV ages and analysis. Need readability. Low Variable ageing &DPI length small and large variance Low readability Projects animals e.g. from Relatively Changes sex. independent good surveys. Need to readability address gear Good selectivity problems growth with dependent function data. possible.

Table 7.9. Summary of monitoring discussions related to datasets use in the workshop. "Yes" means data is useful and "No" is that it is not. 7.5. REFERENCES

Anon. 1996. Queensland subtropical finfish fishery. Queensland Fisheries Management AuthorityDiscussion Paper #3. 50pp. Brown, I.W. Hoyle S., Cosgrove M., Sellin M., Cameron D. and Halliday I. 1996. The South Queensland estuarine/inshore finfish fishery situation Statement #1. ISAMP Project T94/161. 46 pp. Brown IW, Hoyle S., Dichmont C., Sellin M., McLennan M. and Cosgrove M. (in prep). Integrated Stock Assessment and MonitoringProgram (Project No. 94/161). Final Report to Fisheries Research andDevelopment Corporation. Burchmore, J. J., Pollard, D. A., Middleton, M. J., Bell, J. D., and Pease, B. C. 1988. Biology of four species of whiting (Pisces: Sillaginidae) in Botany Bay, New South Wales. Australian Journalof Marine and Freshwater Research, 39: 709-27. Cleland, K. W. 1947. Studies on the economic biology of the sandwhiting (Sillago ciliata C.& V.). Proceedings of the Linnean Societyof New South Wales, 72: 215-28. Dixon, P.I., Crozier, R.H., Black, M. and Church, A. 1987. (Eds). Stock Identificationand Discrimination of CommerciallyImpor tant Whitings in AustralianWaters Using Genetic Criteria (Firta 83/16. 2nd ed. The University of New SouthWales, Sydney. 56 pp. Dredge, M. C. L. 1976. Aspects of the ecology of threeestuarine dwelling fish in south-east Queensland. MSc Thesis. University of Queensland. Gunn, J. S. 1978. A study of the feeding,growth and reproduction of three species of the familySilliganidae in the Townsville region, north Queensland. Honours Thesis. James Cook University, Townsville. Kailola,P.J., Williams, M.J., Stewart, P.C., Reichert, R.E., McNee, A. & Grieve, C. 1993. Australian Fisheries Resources. Bureau of Resource Sciences, Department of Primary Industries and Energy, Canberra.422 pp. Kerby, B.M. & Brown, I.W. 1994. Bream, whiting and flathead in south-east Queensland: a review of the literature. Information Series Q/94028. Department of Primary Industries,30 pp. Maclean, J. L. 1968. A study of the biology of winter whiting Sillago maculata (Quoy andGaimard) in Moreton Bay. MSc Thesis. University of Queensland (Brisbane). McKay, R. J. 1985. A revision of thefishes of the familySillaginidae. Memoirs of the Queensland Museum, 22, 1-73. McKay, R. J. 1992. 'Sillaginid fishes of the world (Family Sillaginidae). An annotated and illustrated catalogue of the sillago, smelt or Indo-Pacificwhiting species known to date', Vol. 14. no. 125. (Pao: Rome-Italy.) Moore, N. 1986. 'Recreational fishing in Hervey Bay and Great Sandy Strait'. no. Ql89021. (Inf. Ser. Dep. Primary Ind. Queensl.: Brisbane, Queensland.) Pollock, B.R. 1980. Surprises in Queensland angling study. Australian Fisheries 39(4): 17-19.

151 Pollock, B. R. 1982. Spawning period and growth of yellowfin bream, Acanthopagrus australis (Giinther), in Moreton Bay, Australia. Journal of Fish Biology, 21: 349-55. Pollock, B. R., and Williams, M. J. 1983. An assessment of the angling fisheryf or summer whiting, Sillago ciliata and S. analis, in Moreton Bay, Queensland from 1959 to 1980. Proceedings of the Royal Society of Queensland, 94: 85-90. Pollock, B. R. 1984a. Management of the fisheryf or yellowfin bream. In 'Focus on Stradbroke. New information on North Stradbroke Island and Newby Islands'. (Eds. R. J. Coleman, J. Covacevich and P. Davie.) pp.348-55. Pollock, B. R. 1984b. Relations between migration, reproduction and nutrition in yellowfinbream Acanthopagrus australis. Marine Ecology Progress Series, 19: 17-23. Pollock, B. R. 1985. The reproductive cycle of yellowfinbream, Acanthopagrus australis (Giinther), with particular reference to protandrous sex inversion. Journal of Fish Biology, 26, 301-11. Quinn, R. H. 1992. 'Fisheries Resources of the Moreton Bay Region'. (Queensland Fish Management Authority:Brisbane.) Steptoe, W. 1995. Advanced flatheadstrategies. Modern Fishing, February 1995 Thwaites, A.J. and Williams,L.E. 1994. The summer whiting fishery in south-east Queensland. Memoirs of the Queensland Museum, 35:249-54. Thwaites, A.J. and Williams, L.E. (unpublished) Review of the yellow-finned bream Acanthopagrus australis (Giinther) fishery in Moreton Bay, Queensland, Australia. Weng, H. T. 1994. Biology of the trumpeter whiting (Sillago maculata) and factors involved in sustainability of stocks in Moreton Bay, Queensland. Ph.D. Thesis, Queensland University of Technology. Williams, L. 1993. Moreton Bay fisheries.In: Moreton Bay in the Balance. pp 71-79. Young, J.A. 1989. Aspects of the reproductive physiology and mariculture of the summer whiting, Sillago ciliata. PhD thesis, University of Queensland

152 SUMMARY OF MAJOR RESEARCH PRIORITIES

The resources reviewed have been treated on a species by species basis, with each species being allocated a chapter in the Workshop Proceedings. At the end of each chapter, a detailed list of monitoring and research directions and priorities are given. The authors present two key recommendations for each species to represent proposals that were given highest priority by the workshop participants in this summary. These recommendations should not be interpreted as representing the exclusive list of proposals and time should be taken to read the full list of recommendations at the end of the report on each species.

153 154 8.1. SPANNER CRAB 8.2. EASTERN KING MONITORING, PRAWN MONITORING, RESEARCH DIRECTIONS RESEARCH DIRECTION AND PRIORITIES AND PRIORITIES Current! y, spanner crabs are monitored In the current management plan, it is solely through the use of commercial proposed to use catch rate levels to act catch rate data. There were many as performance indicators. Participants doubts expressed during the workshop considered that improvement of our concerningthe validity of using simple understanding of the relationship catch rate data as an index of stock between fisheryperformance indicators biomass. The workshop strongly and catch-per-unit-effortdata was a recommended that independent surveys high priority. However, the usefulness of spanner crab biomass were needed of this approach has yet to be demon for long term monitoring in order to strated because catch rates may not be provide an alternative,more defensible closely related to stock abundance. index of abundance. This independent Effortshould be standardised. This survey should commence prior to the project should have a high priority. implementation of the Indivictual With the introduction of Vessel Transferable Quota (ITQ) management (VMS) system, as the integrity of the catch rate Monitoring Systems to the data series may be compromised for a inshore traw 1 fleet,the potential for period thereafter. This research was its use to gather information on fleet given the highest priorityby the dynamics and the distribution of effort workshop. would be of tremendous value in monitoring the status of the stock. Knowledge of spanner crab growth Methods of data capture and sampling rates will greatly enhance Queensland' s intensity should be investigated prior knowledge of possible natural mortality to major investment in softwareand values as well as sustainable catch processes. Theworkshop rated research levels. In the workshop, the value of the in this direction as a high priority NSW tagging data, which relates only because, if successful, VMS methods to adult crabs, was greatly enhanced by could in the long run have a high return dredge samples of juveniles. Although for a low cost. the dredge gear only caught low numbers of juvenile animals, the data improved thegrowth curve estimation. Without informationrelating to the beginning of the growth curve (concerningjuvenile crabs), it is only possible to estimate L= with any precision. Continued dredging for juvenile crabs was considered to be essential by the Workshop. Other methods of ageing crabs should also be investigated.

155 8.3. SAUCER SCALLOP 8.4. SEA MULLET MONITORING, MONITORING, RESEARCH DIRECTION RESEARCH DIRECTION AND PRIORITIES AND PRIORITIES Highest priority was given to the Previous age composition data of the standardized scallop survey. The survey commercial catch appeared to produce conducted in 1997 provided so much the best index of the relative state of the valuable information that it ought to be stock and this age composition data continued, ideally every year. It is, collection should be continued. Due however, a costly project and will to the many age classes, this will be a require financial support from the long-term venture in terms of running industry. Virtual Population Analyses, but the consensus was that the investment Similar to eastern king prawns, the would be highly cost-effective and proposed management plan uses catch beneficial. A non-equilibrium yield rate levels as perlormance indicators. It per-recruit model could be attempted is therefore necessary to improve our in the short term. understanding of the relation between fisheryperlormance indicators and Computer simulations, similar to those catch-per-unit-effort data. Effort completed on spanner crabs within the standardisation will be necessary to workshop, are needed to investigate the improve the relationship between catch Draft Management Plan management rate and biomass. Furthermore, stock rules in terms of trigger frequencyand assessment techniques were most usefulness. These simulations could successful in this session and modelling ultimately be extended to other can be used to support the proposed resources captured by the Ocean Management Plan with tested Beach and Estuarine fisheries. perlormance indicators. This project should have a high priority.

156 8.5. TAILOR 8.6. BREAM, WHITING MONITORING, AND FLATHEAD RESEARCH DIRECTION MONITORING, AND PRIORITIES RESEARCH DIRECTION Validation of the ageing of Tailor is AND PRIORITIES essential to the future assessment and The ageing and length frequency management of this fishery. In the collections of Bream and Flathead workshop it became apparent that the should continue as thisprovides insight state of the stock was showing some into thestatus of these stocks. The indications of stress and overfishing, ageing of whiting was not successful with the degree of overfishing and would not prove as valuable. Over depending entirely upon whether the the long term, forbream and flathead, ageing studies already carried out are ageing informationcould provide corrector biased downward. For a approximate estimates of total mortality small financial outlay a large return in and stock status. management recommendations and resource security could be made. It is necessaryto improve our Another factor that biases the age understanding of the relation between structure collected from the fishery fishery performance indicators and and therefore theestimated fishing catch-per-unit-effortdata, as with mortality, is the theory that large easternking prawn, saucer scallops, and animals move offshoreand are not tailor above. However, in this case, the fished by the shore-based fishers. major stumbling block is the definition Research on the distribution of large of effortin a multi-species fishery and and old fishis thereforehighly the lack of resolution to species of the recommended. In New South Wales catch and effortdata. The fact that these and Western Australia, largefish are species arealso major targets for foundoffshore where spawning occurs recreational fishers,with attendant and this appears to buffer their stocks issues on catch estimates, amplifiesthe fromoverfishing. Queensland tailor difficulties associatedwith assessment appear, at firstsight, not to have this of these species. Research on these natural safeguard. Such research would problems should be given a high not be a trivial undertaking. The priority. offshore distribution of tailor could be determined directly by searchingfor large fishoffshore, perhaps from charter vessel catches. Alternatively, it may be possible to trackonshore schools acoustically to determine whether they move offshoreany great distances.

157 158 AVAILABLE DATA

9.1. INTRODUCTION

Data was collected fromvarious sources, mainly frompast FRDC and other funded projects, QFMA, recreational clubs, Australian National Sportfishing Association and New South Wales Fisheries (FRI). Part of the workshop's success has been gathering this dataset, often offold non-compatible computers or by repunching the data from old datasheets. The full dataset will be kept on CD at Southern Fisheries Centre, Deception Bay and any request to obtain this information should officially be made to Mr Mike Dredge, Industry Manager, SFC, PO Box 76, Deception Bay, 4508. In many cases though, the data are confidential in its present form.

159 160 9.1.1. COMMERCIAL CATCH AND 9.2.1.d CPUE (lifts) by Region and EFFORT DATA Financial yr.xls The commercial catch and effortdata Summarised Catch (t), effort (dilly for all species considered in this lifts) and cpue (kg/lift) by financial workshop were extracted fromthe year and geographic region (within QFMA CFISH database. CFISH is Queensland) fromCFISH data. See a subsystem of QFISH and is fileGRIDREFS.DOC for key to specifically commercialfisheries region codes. data sets. The data originates froma compulsory fisheriescatch and effort 9.2.1.e CPUE (lifts) by Region and daily logbook program that Year.xls commenced in 1988. The data were Summarised Catch (t), effort (dilly retrieved by a 'dilemma' script, lifts) and cpue (kg/lift) by calendar essentially all raw data records for the year and geographic region (within species and dates specified.This was Queensland) fromCFISH data. then put into Access foruse in the workshop. 9.2.1.f Gridrefs.doc Gives geographic limits for 9.1.2. RECREATIONAL CATCH AND assessment regions, and also provides EFFORT DATA a key to the relationship between the Some recreational catch and effort "Historic" grid system and geographic data supplied by QFMA was reg10ns. summarised fromRFISH, a subsystem of QFISH specifically for 9.2.2. BIOLOGICAL AND TAGGING recreational fisheriesdata sets. DATA 9.2.2.a PARAMETERS.doc 9.2. SPANNER CRAB Tabulation of known Ranina DATA population parameters (various sources). 9.2.1. COMMERCIAL CATCH AND EFFORT DATA 9.2.2.b Biotot.xls 9.2.1.a Spanner.mdb Biological datafrom QDPI Raw catch and effortdata for the FIRTASpanner Crab Project Qld spanner crab fisheryfrom 1.1.88 researchcruises between 1981 to about 31 May 1998 extracted from and 1984 inclusive. CFISH. 9.2.2.c Logbook.xls 9.2.1.b Basic Summaryby Year Commercial catch and effortdata (lifts).xls from a voluntary logbook program Summarised Catch (t), effort (dilly run between 1982 and 1987 inclusive. lifts) and cpue (kg/lift) by year across whole Qld fisheryfrom CFISH data. 9.2.2.d Spcrtrip.xls Catch and effort data from QDPI 9.2.1.c Basic Summaryby Year.xls FIRTA Spanner Crab Project research Summarised Catch (t), effort (boat cruises between 1980 and 1984 days) and cpue (kg/boat-day) by year inclusive. across whole Qld fishery from CFISH data.

161 9.2.2.e Sxltot.xls 9.3. EASTERN KING Individual sex and carapace length PRAWN DATA information from researchcruises between 1980 and 1984 inclusive as 9.3.1. COMMERCIAL CATCH AND in Spcrtrip.xls above. EFFORT DATA 9.3.1.a Eastern King Prawn Catch 9.2.2.f Spanner crab length based and Effort data.mdb model specs.doc Raw CFISH catch and effortdata for The specifications forthe length easternking prawns from the based model. TRAWL database. Logbook data retrieval forking prawns only ( other 9.2.2.g Spanner biomass dynamic ° model results.xls species omitted) south of 20 S. Includes other summary (daily, Results of the biomass dynamic monthly) tables and queries. model. 9.2.3. 9.3.1.b Summary of monthly catch NEW SOUTH WALES (NSW) and effort.xls DATA Monthly catch and effortin the 9.2.3.a Co-opdat and co-opgro.x ls Queensland component of the eastern Data from NSW Fisheries Co king prawn fisheryfrom CFISH, operatives, supplied by NSW commencing January 1988 to Fisheries (FRI). December 1997. 9.2.3.b Fecundit and Fecundsu.x ls 9.3.1.c Historical catch and effort data1977-87.xls Fecundity data fromNSW Fisheries (FRI) Research logbook data implemented by Mike Dredge in 1970s. Contains 9.2.3.c Mud.xls about 10% of the easternking prawn N.S.W. Catch and effort statistics vessel records fromthat period, but fromcommercial logbooks. heavily biased to Fraser Island/ Tin Summarised by year. Can Bay Fishery. May be appropriate forhistorical CPUE estimates, but not 9.2.3.d SK4DM.xls total catch. Each large (3-digit) zone N.S.W. Catch and Effort by record includes several of CFISH grids and is with spuriousdata removed. Not to based on the originalCSIRO grid be used to calculate total NSW catch pattern.. Description of blocks as not all catch records are present, available on request. but is excellent forcalculating Catch 9.3.2. BIOLOGICAL AND TAGGING rate. DATA 9.2.3.e Surveyl, Survey2, Sub-adult length frequencies Surveylf.xls Moreton Bay 1988-90.xls Independent survey data fromNSW Monthly length frequency data for Fisheries P. plebejus sampled from nine stations in Moreton Bay over two 9.2.3.f Tagrelea and Tagthing.xls years. N.S.W. tag and release data to estimate growth.

162 9.3.2.a Adultfemale length than the 30' * 30' CFISH grids Details frequency andreproductive data of the grids, based upon a CSIRO 1990-92.xls national grid system, available on Adult femalelength frequency data request. Thedata for 1977-80 fromoffshore waters. Also contains includes approximately 60-70% of the data on ovary weight and histological states scallopcatch. The 1981-87 data condition of 6,000+ individuals from cover a smaller proportion. The 2 data 4 regions (Swain Reefs, Lady Elliot, sets have been combined, using the Mooloolaba and Cape Moreton). larger grids. Samplesobtained each month by the 9.4.1.b QLD-C-F.xls commercial fleetand researchtrawler for reproductive analysis. 1990-92. Summary of catch and effort data fromScallop catch and effort 9.3.2.b Adult male Eastern King data.mdb. Monthly conversion Prawn Length Frequency Data 1990- factors have been applied to convert 92.mdb baskets to kilograms. Adult male obtained in the P. plebejus 9.4.2. BIOLOGICAL AND TAGGING reproductive samples fromthe four DATA regions listed above. Provides monthly length frequency samples of 9.4.2.a 76-77 tag data.xls adult males from the four regions. Monthly tagging program from 76 to 78 for growth and movement 9.3.2.c Tagging data 1990-91.xls information alsoused to estimate Tagging data for9,000+ tagged naturalmortality. 76-77 data only. easternking prawns. Provides growth rate data and also used to estimate 9.4.2.b Bust-growth-est.xls mortality and emigration rates. 9500 tagged scallop release and recapture (2000) details for 5 sites 9.3.2.d Lunar affects.xls for growth variation as a function of Database on lunar and diel variation location and movement, 1991-1992. in reproductive condition and catch rate of adults in offshorewaters, 9.4.2.c Scal-L-W.xls undertaken in 1993. Monthly scallop collection over 12 months in 1977-78 for reproductive 9.4. SCALLOP DATA biology and adductor condition as a functionof time of year and size.

9.4.1. COMMERCIAL CATCH AND 9.4.2.d Survival91.xls EFFORTDATA Subset of Bust-growth-est.xls used to 9.4.1.a Scallop catch and effort estimatemortality as a function of data.mdb exposure time to air after capture. Raw CFISH catch and effort data for saucer scallops from1988-1998 with 9.4.2.e Tag-loss.xls summaryqueries (daily, monthly etc). 526 scallops given multiple tags for Also contains historicalscallop data data on tag shedding rates, 1977 (BH) from1977-87 froma volunteer logbook program operating from 9.4.2.f 88-Bustard.xls Bundaberg, Tin Can Bay, Yeppoon Survey data on numbers and size and Gladstone based on largergrids composition from10'*10'

163 preservation area offBustard Head, 9.5.1.e Total catch.xls December 1988. Yearly catch totals used for graphs 9.4.2.g 89survey.xls within reports. Survey data on numbers and size 9.5.2. BIOLOGICAL AND TAGGING composition fromthree 10'*10' DATA preservation areas, Hervey Bay Y.::,.�.aIll 1�..... u�.J� ....,.. tengtns.xts A'""ll"'lt,'�,.. Yeppoon and Bustard Head in July 1989. Length frequency from 6 sites in the period 1995 to 1996. 9.4.2.h Jan97survey.xls 9.5.2.b 2264 data biologicals.xls Survey data on numbers and size composition fromtwo 10'* 10' Ageing for subsample of length preservation areas, Hervey Bay and frequency. Bustard Head in January 1997. 9.5.2.c Shnute.xls 9.4.2.i oct97survey.xls Shnute models of ageing data. Large scale randomised stratified 9.5.2.d Vonb.xls survey of the main fishing grounds (between 22°30' S and 25°30' S) and Von Bertalanffy models of ageing the 3 preservation areas, Hervey Bay, data. Yeppoon and Bustard Head in 9.5.2.e Regress.xls October 1997. Regression of Total Length vs Fork 9.5. MULLET DATA Length, Wet Weight vs FL etc. 9.5.2.f Estimating growth from 9.5.1. COMMERCIAL CATCH AND papers.xls EFFORT DATA Published and unpublished growth 9.5.1.a Mullet.mdb curves. Raw mullet catch and effortdata from 9.5.2.g Male and female GSI.xls 1988 to 1997 extracted from CFISH. Monthly GSI for 1995 and 1996. 9.5.1.b Lew's Tot @ OBF vs non OBF.doc 9.5.2.h Juvdat.xls Summary of commercial catches by Small amount of data on juvenile fish region and ocean beach fishery for age and lengths. catches vs estuarine catches. 9.5.1.c Historical catch H&M 9.6. TAILOR DATA reports.docs 9.6.1. COMMERCIAL CATCH AND Historical catches fromHarbours and EFFORT DATA Marineannual reports. 9.6.1.a Historic.xls 9.5.1.d Com catch by month.xls Data fromthe Queensland Fish Board Commercial catches by month by dating back to 1944. The Fish Board reg10n. operated until 1981. Data are from processor's returns. There was probably a large 'black market' component not recorded here. The

164 data from 1987 on come fromthe 9.6.1.i Aldo's beach and rock QFMA's CFISH database. data.xls 9.6.1.b Cpue etc.xls Compares the fork lengths of beach, rock, and breakwall caught tailor.The Summary of CFISH commercial data were provided by Aldo Steffeof logbook data for tailor from 1988-97. New South WalesFisheries, froma 9.6.1.c Recreational Catch and creel survey of recreational fishers. Effort data 9.6.lJ Aldo's boat-based data.xls 9.6.1.d Fraser permits DOE.xls Compares tailor catch from boat and shore based anglers. Results froma Data from access permits issued by GLM analysis are included. The data the Departmentof the Environment. were provided by Aldo Steffeof New This is an attempt to look at tailor South WalesFisheries, froma creel fishing effort. However, Fraser Island survey of recreational fishers. is only a small part of the Queensland and NSW tailor fishery,and many 9.6.2. BIOLOGICAL AND TAGGING people obtaining permits for Fraser DATA Island go there as tourists, not as 9.6.2.a St961105.xls tailor fishers. Shows dependency of fork lengthon 9.6.1.e All tailor club data.xls 'catch', where a catch is a number of fish takenby a group of anglers in a Data recorded as part of club single location during a morningor competitions. The data were supplied evening fishing session.Indicates by QFMA, who obtained it from tailors' tendency to school in groups Queensland fishingclubs. Included on of similar size. the sheet is a graph of the average weight of tailor caught. These show 9.6.2.b Not all data.xls an increase after 1986. A SAS GLManalysis of theweight 9.6.1.J Location codes.xls trends in the 'All tailor club data.xls'data. The 'not all data' refers A list of thelocation codes used in to theomission of most of the catches, the All tailor club data.xls. where no weights were recorded. 9.6.1.g Fishing club-all surf beach 9.6.2.c Barry Pollock's tagging tailor.xls data.xls Surf beach tailor data only from These data come froma tagging All tailor club data.xls. program carriedout from 1978 to 9.6.1.h Beach out.sas, 1980. Recaptures are also entered Beach.pgm.sas, Beach.sd2 on the sheet. SAS files for GLM analysis of catch 9.6.2.d Pepperell.xls numbers (not weight) on ocean The data come froma tag recapture beaches, data fromFishing club-all program along much of the NSW surf beach tailor.xls coastline which began in July 1976, occurred intermittentlyuntil 1980, then more regularly until 1982. Provided by Julian Pepperell,

165 formerly of New SouthWales 9.6.2.l Tailor growth 98-06-25.xls Fisheries. Growth curves calculated from 9.6.2.e Tag data 87-88.xls, tag data Integrated Stock Assessment Project 89.xls, recaps for 87-88.xls, recap data. 89.xls, Halliday's data-mixed.xls, Halliday's-Tag89.xls, Halliday's 9.7. BREAM, WHITING Tag90.xls. AND FLATHEAD DATA These files come froma DPI Tailor 9.7.1. COMMERCIAL CATCH AND Tagging project undertaken from EFFORT DATA 1987 to 1990. The fish were caught, measured, and tagged by a group of 9.7.1.a Isamp species.mdb club fishers. Raw CFISH catch and effort data for 9.6.2.f Ag961111- Z estsfromall bream, flathead and whiting fromthe lengths.xls Mixed database. Tables for each species group and queries for Z estimates fromlength data from summarising data. Fraser Island, using the equation Z=K(Linf-mean(L))/(mean(L)-Lcrit). 9. 7.1.b Historic.xls 9.6.2.g AG961113-Zfrom rec, co m Summarised data from the Queens lengths.xls land Fish Board dating back to 1944. The Fish Boardoperated until 1981. Separate estimates of Z from the all Data are fromprocessor's returns. the available recreational and There was probably a large'black commercial sampling data to see if market' component not recorded here. presumed size selectivity differences The data from 1987 onwards are affect the estimates. derivedfrom the CFISH database. 9.6.2.h Edited with extras2.xls 9.7.2. RECREATIONAL CATCH AND Tailor marginal increment analyses EFFORTDATA showing that otolith rings form in the 9.7.2.a Sample size to observe latter partof the year, fromAugust to declining 15% catch rate.xls December. Analysis of 1997 recreational fisher 9.6.2.i All von B curves.xls survey on bream and whiting in Pumicestone Passage. Looking at the A number of von Bertalanffy growth option of monitoring catch rates using curves devised for tailor around the recreational fisher surveys. world are set out here, including two fromTerry Bade in Queensland (1980 9.7.2.b Sample size to observe MSc thesis). declining catch rate.xls 9.6.2.j Bluefish length wt.xls As above with a 10% decline rather than 15%. Bluefish length-weightrelationship fromWilk 1977. 9.7.2.c CS recfish total lengths.xls 9.6.2.k Growth and comparisons of Length frequencydata for bream and size by catch.xls summer whiting from Coastal Streams Project. Data from 3 regions - Length frequenciesfrom Integrated Maroochy River, Burnett River and Stock Assessment Project data. Pumicestone Passage. 1997.

166 9.7.2.d CS spring 1997 survey 9.7.3.c Bream marginal increments catches br, flat, tail, whit.xls analysis (prelim).xls Recreational fisher survey data from Marginal increment data and analysis Coastal Streams Project spring1997 forbream fromthe Integrated Stock for Maroochy River, BurnettRiver Assessment Project. and Pumicestone Passage. Data on bream, flathead, tailor and whiting. 9. 7.3.d Growth rate - rough estimate.xls 9.7.2.e CS winter 1997 survey Ages 1 to 4 omitted because of the catches br, flat, tail, whit.xls influenceof the legal size at this level. Recreational fishersurvey data from The O age fish come from research Coastal Streams Project winter 1997 data. >4 age data fromcommercial for Maroochy River, BurnettRiver catch sampling. and PurnicestonePassage. Data on bream, flathead, tailor and whiting. 9.7.3.e Size varies by sex, catch, and gear type.xls 9.7.2.f Pumices tone 93 catch.xls A SAS GLM analysis of size Pumicestone Passage recreational selectivity of catches (fishfrom fishersurvey, 1993. (individual fish 1 commercial fisher on one day). data) Lengthvaries by sex, gear type and catch. 9.7.2.g Pumicestone93 creelsurve y.xls Pumicestone Passage recreational 9. 7.3J Growth, M_F compare.xls fishersurvey, 1993. (individual boat Estimates of von Bertalanffy growth data) rates fromIntegrated Stock Assessment Project flatheaddata. 9. 7.2.h Jumpinpin bream club Comparisons of male and female data.xls growth rates in Moreton and Yearly summaries taken from an Hervey Bays. RFISH historical fishing club database. 9.7.3.g Flatheadlengths.xls Lengths from commercial flathead 9.7.2.i Jumpinpin whiting club catch 1991-93. data.xls Yearly summaries takenfrom an 9.7.3.h Flatage.xls RFISH historical fishingclub Ages estimated fromotoliths for database. flathead fromcommercial, recreational and scientificsamples, 9.7.3. BIOLOGICAL AND TAGGING DATA 1991-93. 9.7.3.a Fish97.mdb 9.7.3.i Growth boot- Camoe's pumicestn data.xls The Integrated Stock Assessment Project 1995-1998 database. All the Estimation of growth rates from project's data are stored in here. Pumicestone Passage recreational fish survey, 1991-1993. 9.7.3.b Bream.mdb Data purchased fromthe ANSA sportfishingprogram, club fisherstag and recapture details.

167 168 ACKNOWLEDGEMENTS

The Fisheries Research and Development Corporation (FRDC), Queensland Department of Primary Industries (QDPI) and QueenslandFisheries Management Authority (QFMA) funded this workshop. It was a major undertaking for several people, however, the largest contributionscame from Dr Malcolm Haddon, workshop facilitator, Proceedings author and editor; Ms Kath Kelly, workshop coordinator and transcriptor of the audio tapes; Ms Kate Yeomans, data collector, collator and Proceedings editor. Your dedication is much appreciated. I would like to thank DrIan Brown and Mr Bob Pearson for convincing the various parties andinterest groups that this workshop was essential and for securingfinancing from QFMA and QDPI. Each session had an allocated coordinator, which really reduced my workload. They were; Dr Ian Brown for spanner crabs, Dr Tony Courtney for easternking prawns, Mr Mike Dredge for scallops, Mr Ian Halliday for mullet and Mr Simon Hoyle for tailor and the bream/ whiting/flathead sessions. Several scientists have reviewed the manuscript, but I would like to make a special mention of Mr Mike Dredge, Dr Ian Brown and Mr Daryl McPhee. I would also like to thank the laboratory for allowing me to use the tearoom and other facilitieswithout complaint. Most of all, I really appreciate the help from the administrative stafff or what was a difficult and diverse project.

169 170 PARTICIPANT LIST

1. Mr David Bateman, Sunfish, 10 Dalloon St, Boondal, QLD, 4034. 2. Mr JeffBibby, QFMA, Brisbane, Queensland Fisheries Management Authority, Box 344, Fortitude Valley, 4006, [email protected] 3. Mr JeffBlaney, Fisher, Contactable through QCFO, PO Box 392, Clayfield, QLD, 4001. 4. Dr Ian Brown, DPI, Deception Bay, Department of PrimaryIndustries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] .au 5. Mr Rik Duckworth, NT Fisheries,Northern Territory Department of Primary Industry and Fisheries,GPO Box 990, Darwin, NT, 0810, [email protected] 6. Dr Charis Burridge, CSIRO, Cleveland, Division of Marine Research, CSIRO, PO Box 120, Cleveland, QLD, 4163, [email protected] 7. Mr Adam Butcher, DPI, Deception Bay, Department of PrimaryIndustries, Queensland, Southern FisheriesCentre, Box 76, Deception Bay, QLD, 4508, [email protected] 8. Mr Darren Cameron, QFMA, Brisbane, Queensland Fisheries Management Authority, Box 344, Fortitude Valley, 4006, 9. Mr Michael Cosgrove, DPI, Deception Bay, Department of Primary Industries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] 10. Dr Tony Courtney, DPI, Deception Bay, Department of PrimaryIndustries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] 11. Ms Cathy Dichmont, DPI, Deception Bay, Department of PrimaryIndustries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] 12. Dr David Die, CSIRO, Cleveland, Division of MarineResearch, CSIRO, PO Box 120, Cleveland, QLD, 4163, [email protected] 13. Mr Mark Doohan, QFMA, Brisbane, Queensland Fisheries Management Authority, Box 344, Fortitude Valley, 4006, [email protected] 14. Mr Russel Doyle, Fisher, 86 Lapoinya Cres, Warana, QLD, 4575. 15. Mr Mike Dredge, DPI, Deception Bay, Departmentof PrimaryIndustries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] 16. Dr Malcolm Dunning, DPI, Brisbane, Department of Primary Industries, Queensland, 9th Floor, Forestry House, 160 Mary St, Brisbane, QLD, 4001, [email protected] .au 17. Mr Brendan Farthing, QUT, Brisbane, Queensland University of Technology, Garden Point, Brisbane,8000, [email protected] 18. Mr Richard Freeman, Fisher, 11 Robe St, Currimundi, QLD, 4551. 19. Mr Peter Gaddes, Fisher, 3 France Court, Newport, QLD, 4020. 20. Mr Richard Gilbert, Fisher, 37 Swan Drive Bundaberg, Bundaberg, 4670.

171 21. Mr Dale Golian, NSW Fisheries, New South Wales Fisheries, PO Box 21, Cronulla, NSW, [email protected] 22. Dr Charles Gray, NSW Fisheries, New South Wales Fisheries Institute, PO Box 21, Cronulla, NSW, 2230, [email protected] 23. Dr Neil Gribble, DPI, Cairns, Department of Primary Industries, Queensland, NorthernFisheries Centre, Box 5396, Cairns,4870, [email protected] 24. Dr Malcolm Haddon, AMC, Head, Department of Fisheries, Australian :Maritime College, P.O. Box 21, Bcaconsficld, Tasmania, 7270, [email protected] 25. Mr Ian Halliday, DPI, Deception Bay, Southern Fisheries Centre, Box 76, Deception Bay, 4508, [email protected] 26. Mr Jim Higgs, QFMA, Brisbane, Queensland Fisheries Management Authority, Box 344, FortitudeValley, 4006, [email protected] 27. Dr Burke Hill, CSIRO, Cleveland, Division of Marine Research, CSIRO, PO Box 120, Cleveland, QLD, 4163, [email protected] 28. Mr Simon Hoyle, DPI, Deception Bay, Department of Primary Industries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] .au 29. Mr Eddie Jebreen, DPI, Deception Bay, Department of Primary Industries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] 30. Ms Kath Kelly, DPI, Deception Bay, Department of PrimaryIndustries, Queensland, Southern Fisheries Centre, Box 76, DeceptionBay, QLD, 4508, [email protected] 31. Dr Steve Kennelly, NSW Fisheries, New South Wales Fisheries Institute, PO Box 21, Cronulla,NSW, 2230, [email protected] 32. Dr Don Kinsey, Finfish MAC Chair, Lodge Rd, Mountain Top, Via Nimbin, 2480, [email protected] 33. Dr John Kirkwood, DPI, Deception Bay, Department of PrimaryIndustries, Queensland, Southern Fisheries Centre,Box 76, Deception Bay, QLD, 4508, [email protected] 34. Mr Vern Lee, Trawl Fisher and Processor.Lee Fishing Co Pty Ltd, PO Box 54, Tin Can Bay, QLD, 4580. 35. Dr David Mayer, Biometrician, Animal Research Institute, 665 Fairfield Road, Yeerongpilly, 4105, [email protected] 36. Mr Mark McLennan, DPI, Deception Bay, Department of Primary Industries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] 37. Mr Daryl McPhee, QCFO, PO Box 392, Clayfield, QLD , 4001, [email protected] 38. Mr Dave Mitchell, Crab MAC Chair, 28 Bailey Ave, Taragindi, QLD, 4121. 39. Mr Mick O'Neill, DPI, Deception Bay, Department of PrimaryIndustries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] 40. Dr Jenny Ovenden , DPI, Deception Bay, Department of Primary Industries, Queensland, Southern Fisheries Centre, Box 76, Deception Bay, QLD, 4508, [email protected] .au 41. Dr Ian Poiner, CSIRO Cleveland, CSIRO, Box 120, Cleveland, 4163, [email protected] 42. Dr Barry Pollock, DPI, Brisbane, Department of PrimaryIndustries, Queensland, 5th Floor, Forestry House, 160 Mary Str, 4001, [email protected] 43. Dr Andre Punt, CSIRO Hobart, CSIRO Division of Marine Research, GPO Box 1538, Hobart, Tas, 7001, [email protected]

172 44. Dr James Scandol, QTUF,Quantatative Training Unit for Fisheries,Marine Ecology Laboratories,Univ. Sydney,NSW, 2006,[email protected] 45. Mr Peter Seib, Fisher,31 Sippy Downs Drive, Sippy Downs,4556. 46. Ms Michelle Sellin, DPI, Deception Bay,Department of Primary Industries, Queensland, Southern Fisheries Centre,Box 76,Deception Bay,QLD, 4508, [email protected] 47. Dr Ian Smith, QFMA,Brisbane, Queensland Fisheries Management Authority, Box 344,Fortitude Valley,4006, [email protected] 48. Mr Jonathan Staunton-Smith, DPI,Deception Bay, Department of Primary Industries,Queensl and, Southern FisheriesCentre, Box 76, Deception Bay, QLD,4508, [email protected] 49. Mr Richard Steck.is,WA Fisheries, WesternAustralia Marine Research Laboratories,Research Services Division,Fisheries Department of WA, PO Box 20,North Beach,WA, 6020,[email protected] 50. Mr Dave Sterling, Fisher,DJ Sterling Trawl Gear Services,12 Oyster Parade, Tin Can Bay, QLD,4580, djstgs.tpgi.com.au 51. Mr Wayne Sumpton, DPI, Deception Bay, Departmentof Primary Industries, Queensland, SouthernFisheries Centre,Box 76, Deception Bay,QLD, 4508, [email protected] 52. Mr Clive Turnbull, DPI, Cairns,Depar tment of Primary Industries, Queensland,Northern Fisheries Centre,Box 5396,Cairns, 4870, [email protected] 53. Mr John Virgona, NSW Fisheries,New South Wales Fisheries Institute, PO Box 21,Cronulla, NSW, 2230, virgonaj@fishe ries.nsw.gov.au 54. Dr Reg Watson, WA Fisheries,Western Austral ia Marine Research Labroatories,Research Services Division,Fisheries Department of WA,PO Box 20,North Beach,WA, 6020,[email protected] a.gov.au 55. Mr Lew Williams, DPI,Brisbane, Department of Primary Industries, Queensland,9th F1oor, Forestry House,160 Mary Str.,Brisbane, QLD, 4001, [email protected] 56. Ms Kate Yeomans, DPI,Deception Bay,Department of Primary Industries, Queensland, Southern Fisheries Centre,Box 76, Deception Bay,QLD, 4508, [email protected]

173 174 COMPUTER SOFTWARE UTILISED

• SAS/STAT & SAS/BASE. Statistical Analysis Software v.6.12. SAS Institute Inc. North Carolina, USA.

• Arcview GIS v3.1. ESRI, Redlands, California, USA

• Genstat. V5 R4.1 The Numerical Algorithms Group Ltd., Oxford, UK.

• Microsoft Office97, Microsoft, Redmond, WA, USA

• Ad Model Builder, Otter Research Ltd., Nanaimo, British Columbia, Canada.

175 176 ABBREVIATIONS

AMC Australian MaritimeCollege CFISH Commercial Fisheries Information System (a subsystem of QFISH) CL Carapace Length CPUE Catch per unit effort Crab MAC Crab Management Advisory Committee Crab SAG Crab Stock Assessment Group CSIRO Commonwealth Scientificand Industrial Research Organisation CV Coefficient of Variation DPI Department of Primary Industries, Queensland DPIF Departmentof Primary Industries and Fisheries, Northern Territory F Fishing Mortality rate GLM General linear model ITQ Individual transferable quota NSW Fisheries New South Wales Fisheries NSW New South Wales Ppt Parts per Trillion QFB Queensland Fish Board QFISH Queensland Fisheries Information System QFMA Queensland Fisheries Management Authority QUT Queensland University of Technology RFISH Recreational Fisheries Information System (a subsystem of QFISH) TAC Total allowable catch TL Total Length WA Western Australia YPR Yield-per-recruit

177 178 TRANSCRIPTS

The Workshop transcriptsare on the disk attached to thisProceedings. The transcriptsare fromdiscussions duringand afterthe seminar sessions and during the monitoring discussions. Contents have been included in the files. Below are the appropriate filenames:

14.1. DISCUSSIONS DURING SEMINARS • Spanner crab discussions during talk.doc • Easternking prawn discussions during talk.doc • Saucer scallop discussions during talk.doc • Mullet discussions during talk.doc • Tailor discussions duringtalk.doc • Bream, whiting and flathead discussions during talk.doc

14.2. DISCUSSIONS ON MONITORING ANDRESEARCH DIRECTIONS • Spanner crab monitoring discussions.doc • Easternking prawns monitoring discussions.doc • Saucer scallop monitoring discussions.doc • Mullet monitoring discussions .doc • Tailor monitoring discussions.doc • Bream, whiting and flatheadmonitoring discussions.doc

179