P2301 - ICES
Contents
1. i L Weghing Methad ey Weighing Method ame OF Weighiing Method Figure 5 Main menu of VPABase E VPABase Version 0 1 19 July 2001 MAIN MENU Enter data on screen Enter data from EXCEL worksheets Select samples for unsampled Catch and process data View Inventory and List of samples ACCESS reports View Reports in Excel Exit s e mle Lx E 24 Figure 6 Data entry form of VPABase amp catch Data entry E ass VPABase ANNUAL CATCH BY COUNTRY Country Netherlands Official Landings 459 tons Year 2001 Unreported Landings 75 us loxi Save and Exit Species Cod E MisAllocated Landings 82 ICES Sub Area Iv x Discards ICES Division Ve E Sampled Catch be Add new ICES Sub Division All z record CATCH BY FLEET AND TIME PERIOD Quarter 101 Jan Feb Mar Use Of Catch Fleet Number of samples Stock North Sea Cod Weight of Catch Date of Recording AGE COMPOSITION SAMPLE W rings Catch No O00 Body weight kg No Lengths M 880 233 Record i 4 I 1 gt or rx of 4 Record i4 4 J 1 rijp of 12 25 Figure 7 Menu for data processing amp 3 Create Allocation table E Dl x EE VPABase MENU FOR DATA PROCESSING i aS Create new allocation table 2 E Select samples for unsampled Catch 3 Save selection of samples 4 n Back to main menu Figure 8 Data entry form for the selection of age com2. nl Q a number of length measurements CVc Q a Coefficient of variation of the estimate of Catch Q a CV wc Q a Coefficient of variation of the estimate of wc Q a 3 2 Sampled vs unsampled strata A sampled stratum 1s a stratum in which biological sampling was conducted The catch numbers at age and the weights at age could be directly estimated The total yield is also available An un sampled stratum is a stratum where only the total yield is known and no sampling data are available The estimation of age composition and weight at age for un sampled strata has thus to be done using information from the sampled strata The set of all strata may be split into two the sampled strata S and the un sampled strata S sample The number of sampled and unsampled strata are noted mg and my respectively A sampled and unsampled stratum is noted QS and QY respectively Thus the set of all strata can be partitioned as U indicates the union of two sets S Sampled S S S Un Sampled U U U SSS saa uisi e NE NE a Oso s U In a sampled stratum 2 the total number caught of all age groups is amax St Catch Q e Catch 2 a 3 a A min St The proportion of each age group in the total catch number in the stratum is Preach 2 a Catch 2 a Catch 2 e 4 The mean weight of all fish in the catch is ama St We QS V Proun 25 2 W 2 a 5 a Min St The computed total weight of th
3. However VPABase is at this stage of implementation only a prototype that requires both improvements and testing before being considered as possible standard software in ICES The first improvement would be the ability to output not only the matrices of catches and weight at age but also the tuning files That would require extending the fleet record level to account for effort data and also integrating survey data The second improvement would be to develop VPABase into a fully integrated software that would perform the whole assessment procedure instead of producing only input files to be used in existing software That would avoid all problems due to format specifications and changes from one software to another as it 1s currently the case in ICES working groups The usual software used in fish stock assessment could be rewritten in a module in VPABase e g in Visual Basic which would become a full assessment package Here again the advantage of using ACCESS is the possibility for the user to see and modify the programming code and thus to use the database and assessment programs with high flexibility and transparency The third improvement would be to add a plat form independent module to the database and to improve its access and use Although ACCESS is widely used by a number of institutions problems arise when using another operating system or when updating the database then a replica has to be sent to each individual partner each
4. ICES as the Official statistics These statistics are however sometimes modified by members of assessment working groups This section discusses the discrepancies between official statistics and the so called working group data as all the data included in the working group data have to be included in the database 4 1 Official statistics and working group data Let Landings Woffcia St Ct Ar be the official statistics i e the weight of the annual landings of stock St by country Ct in ICES area sub area or division Ar Members of ICES working groups deliver the so called working group data which may be more disaggregated than the official data and which may also be different The question is whether the official data are consistent with the working group data that is Fly St 4 LandingsWog StCt Ar gt M Yield St FI Ar q 13 Fl 1 AreICES Div q 1 Reasons for discrepancies might include the non integration of discards data as well as corrections made on misreported data The following fields are used by the Working Groups ICES 2000 and are integrated in the VPABase Official Catch Catches as reported by the official statistics to ICES Unallocated Adjustments to the official catches made for any special knowledge Catch about the fishery such as under or over reporting for which there is firm external evidence can be negative Area Used only to adjust official catches which have
5. Proto type of fleet record for data exchange FLEET RECORD Can be all Fleets or set of fleets Given only if Div or sub div not filled in Division Given only if sub div not filled in Sub division 29135 when SOP available this field need not be filled in 38692 when SOP available this field need not be filled in Total number Average Body Weight Coefficient of landed discarded Number of Otoliths during period Variation Discard Age 1 Discard Age 2 Discard Age 3 Discard Age 4 20 11 FIGURES Figure 1 Illustration of the gap filling process The borrowing of samples is illustrated in the figure by the broken lines from a sample to a catch age composition FLEET A COUNTRY A FOR FLEET B gt COUNTRY A 3 FLEETC X COUNTRY A dns eum FLEETA Pe COUNTRY B EXPLANTION ut COUNTRY Age compostion data B total numbers F caught by age group COUNTRY B mple c Pa 25 Sample used for estimation but not taken from catch i I i Sample takem from catch Figure 2 The basic hierarchy of tables in VPABase FLEET Age Comp RECORD record Age Comp FLEET record RECORD COU NTRY Age Comp record RECORD FLEET Age Comp RECORD record FLEET RECORD 21 Figure 3 Data Tables and relations of the ACCESS implementation of VPA Base Compare Table 1 Note that the two tables T Use Of Samples and T Use Of Samples Saved are the tables containin
6. the yields weight of catches The three principal data tables of VPABase are 1 Country record 2 Fleet record 3 Age composition record In addition there is the auxiliary data tables containing information on the data processing 4 All combinations of catch and samples for estimation of age composition 5 Selected samples for estimation of age composition The tables are linked in a hierarchy of one to many relationships Table 2 shows the fields in the input data tables and the auxiliary tables The first fields in the tables e g Key Country Catch in the table T Country Catch are the keys linking the tables in the hierarchy Table 2 indicates the relationships between the tables which is also illustrated by Figyre 3 which by arrows indicates the relations of the three principal data tables and the auxiliary tables Figure 4 shows the complete set of look up tables together with the three principal input data tables A complete description of all fields keys and relations of VPABase is rather voluminous The reader interested in further details is referred to the reference manual of VPABase The auxiliary tables T Use of Samples and T Use Of samples Saved contain the input which specifies the processing of data It contains information on which sampled records were used to estimate age compositions of the un sampled records They also contains information on the methodology of data processing as well as the time
7. 27
8. International Catch at age and Weight at Age 9 LIST OF SYMBOLS a Age group ayax St Oldest age group of stock St which may be a plus group aui St Youngest age group of stock St gt 0 Ar Area index Ar 1 2 Atyax ATMax Aryax Number of areas Catch 2 e The total number caught of all age groups in stratum Q Catch Q a Catch Landings discards in numbers of age a in stratum Q Ct Country CVc Q a Coefficient of variation of the estimate of Catch Q a CV c Q a Coefficient of variation of the estimate of W Q a Cypa St yq a Input to VPA Working group data DiscardN Q a Numbers discarded in stratum Q age group a Working group data DiscardW Q a Yield discarded in stratum Q Working group data Fl Fleet index Fl 12 Fly Flu Ct Flya Number of fleets of country Ct LandingsN Q a Numbers landed in stratum Q age group a Working group data LandingsW Q a Yield landed in stratum Q age group a Working group data LandingsW opficia St Ct Ar Weight of annual landings by country Ct Official statistics delivered from governments to ICES ms Number of sampled strata M Method of weighting sampled strata to estimate unsampled distributions mu Number of un sampled strata n Q a Number of age measurements otoliths n Q a Number of length measurements Pr catch Q a The
9. Q a Numbers discarded 5 Yield Landings or discards Yield Q LandingsW Q or DiscardsW Q Weight of catch Landings or discards 6 Number of otoliths nc Q a Problematic see below Inverse CV Age 1 CVc Q a Only for catch at age e g from Bootstrap 8 Inverse CV weight 1 CVwc Q a Only for weight at age e g from Bootstrap Table 2 Fields in input data tables of VPABase COUNTRY CATCH RECORD Name in ACCESS T Country Catch see Fig 3 Name of Field nique key of country record Look up name of country Look up A species can be separated into Stocks CES Sub division CES Division CES Area Official Landings nreported landings IMisallocated landings 1999 2000 2001 Look up name of ICES subdivision Look up Not filled in if above filled in Look up Not filled in if above filled in As reported to ICES by governments WG members estimate of unreported landings WG members estimate of landings WG members estimate of discards The weight of the catch covered by samples Continued 16 Table 3 Continued Fields in input data tables of VPABase FLEET CATCH RECORD Name in ACCESS T Fleet Catch see Fig 3 nique key of fleet record Also Pointer to sample if existing which was taken fro his catch ointer to country record one to many Part of country where home ports of fleets are optionally Look up Look up N
10. been reported from the misreported wrong area can be negative For any country the sum of all the area Catch misreported catches should be zero Discarded Catch Catch which is discarded WG Catch The sum of the 4 categories above 4 2 Discards and landings Estimated catches in numbers and yield in weight is composed of landings and discards Catch Q a LandingsN Q a DiscardN Q a Yield Q LandingsW Q DiscardW Q The weight at age of discards and landings are usually different the discard weight age age being smaller than the landings weight at age Wp Q a average body weight of discards W Q a average body weight of landings The body weights relate to the body weight at age in the catch by W Q a LandingsN Q a Wp Q a DiscardsN Q a W Q a cl a LandingsN Q a DiscardsN Q a 14 whereby ama St Landings W Q gt Ww Q a LandingsN Q a a a yin St and ama St Discards W Q 25 Wp Q a DiscardsN Q a d yi St 4 3 Misreported and unallocated catch When including the terms for unallocated and misreported landings as defined in the text table above the following equations should always hold Landings W oficia St Ct Ar UnallocatedW St Ct Ar CatchW St Ct Ar 15 MisreportedW St Ct Ar DiscardW St Ct Ar where 3 Misreported St Ct Ar 2 0 AreICES Div and 4 CatchW St Ct Ar gt gt Y
11. by the ICES Herring WG and the Mackerel Horse mackerel Sardine and ANCHOVY WG i ee mibi mete He EO d e hiis t ced ee aaa 27 Figure 10 Example of report in EXCEL format sese reete rennen 27 1 Introduction Biological advice on many commercially exploited stocks in the ICES area relies to a large extent on commercial catch at age The quality of the biological advice is dependent on the quality of the underlying data However for most stocks there is hardly any insight into measures of uncertainty associated with the catch at age data Furthermore for the stocks exploited by various countries the sampling and raising procedures to estimate national age composition of catches might differ from one country to another The methods used for combining national catch statistics into international catch at age are not uniform The historical disaggregated national catch data are not always accessible and thus the consistency of data series is not warranted Thus the whole procedure of estimating international catch at age would gain in efficiency and transparency if a uniform method of aggregation was developed and applied and if all data used in the estimation were gathered and centralised in one single database The objective of this paper is to present a prototype of such a database for the storage and the processing of ICES Working Group input for fish stock assessment Input data are total catches by fleet and country area and t
12. fey old Record i 4 I gt Leif of 37 26 Figure 9 EXCEL worksheet used by the ICES Herring WG and the Mackerel Horse mackerel Sardine and Anchovy WG The figure shows only a part of the worksheet ene esi ze ye pe t et e d e d pe pe ee ee em e d M N 0 CATCH NUMBERS LENGTH AND WEIGHT AT Catch data VG estimate should include discards and unre WG for the Assessment of Herring in the Area south of 62 TONNES SHOULD ALWAYS BE FILLED IN even if no numbe SI 2 Eam 4 pema o Division Vaw 6 veg Revised date 3 mar 00 only t 7 i 8 3 Numbers Mean H i Mean Mumbers Mean Mean Mumbers Mean Winter at age d Length Wei atage i Length Weight atage Length d weight atage Length Weight Rings 000 em ka 000 em kg 5000 fem i 30 New sheets for other fishing areas may be copied by EDIT MOVE OR COPY SHEET Then rename the sheet according the new fishing area by double clicking on it Rename also this CANUM sheet ae gt Mf CANUM IvaE CON CANUM IVaW CON CANUM IVaE IND CANUM IVaW IND CANUM IVB IND i Bir Figure 10 Example of report in EXCEL format Ed Microsoft Excel vpabase output xls File Edit View Insert Format Tools Data Window Help D SH S 6RY mmo o c J amp x hr 4 Winter rings Numbers 000 Mean weight D 10175 2116 0 CO 4 Co C1 ER Wh
13. for unsampled catch The userform for selction of samples is shown in Figure 8 The upper left part of this form contains a number of fields headed Catch record showing the particulars of the catch record to be processed These fields are shown only for information The fields are locked so that the data cannot be entered or modified through this form There is a field in the centre of the form which in the example of Figure 8 contains the information No age distribution for this catch hery Weight of catch written in red This means f Weight of catch n NO age distribution that the catch of the current for this catch record is un sampled and you need to select samples for estimation of age compo sition If there is a an age be empty composition sample the field w1 The form has a check field in the right column to indicate which sample is selected for a given catch presented in the top portion of the form The user can change the selection by removing or adding check marks on the form 11 In the hypothetical example of Figure 8 two samples no 138 and 139 originating from United States industrial fishery is used to represent Danish industrial fishery In the upper right hand corner of the form it is possible to change weighting method for pooling of samples In the example of Figure 8 the samples have been weighted by the Sampled catch The weighting methods only applies if more than one samp
14. later processing of samples The data entered through look up tables appear as list See oe boxes in the user form When clicking on the right hand side Quarter Q1 Jan Feb Mar arrow button a list of options is displayed from which you Fleet OB Trawl can select the input The example given here is taken from Stock North Sea Cot Figure 6 showing the entry of Stock Only input from the list is accepted Only users with special privileges can change the content of the list box This approach should reduce the risk of mistyping and ambiguities in data North Sea Cod Via Cod Not specified ee In later versions of VPABase the data entry form will contain a number of additional facilities such as production of graphs and tables as well as facilities for finding records and for deleting records Later versions of VPABase will also be able to handle catch data by statistical rectangle Entry of tuning data in future versions of VPABase will probably be done in a separate userform allowing for entry of catch and effort or CPUE 10 5 4 Automatic data Input The second item on the main menu lets the user import data from a predefined EXCEL format or a text file see example of EXCEL input in Figure 9 The present version of VPABase can read the EXCEL sheets currently used by the Herring WG and the Mackerel Horse mackerel Sardine and Anchovy WG ICES 2000 In later version
15. of last processing This part of the database is perhaps the most complicated part of it as the number of choices for combination of sample and processing method 1s usually large A major objective of VPABase is to store the information on data processing in an unambiguous way 5 2 User interface The purpose of the user interface is to guide the user through three steps 1 Input data 2 Make decisions on estimation of age compositions for un sampled strata by selecting age compositions among samples strata 3 Generate output tables The user interface implemented through user forms that are of two types 1 Menus selection of options 2 Data entry form The five first items of the main menu Figure 5 guides the user through the different steps to produce the aggregated input data for VPA Options 1 and 2 are alternative options namely 1 to Enter data from the keyboard se next section or 2 Enter data from an EXCEL worksheet see Section 5 4 Item 1 in the main menu ends with which indicates that this option leads to another user form see Figure 6 Item 3 activates the user form for data processing Items 5 and 6 deal with the reports that is the output Output can be in ACCESS format Item 4 or in EXCEL format Item 5 In later versions of VPABase the main menu will contain an option for entry of tuning data survey data commercial fishery effort data or commercial fishery CPUE 5 3 Manual d
16. proportion made of age group in the total number Pr on Q S Mt Weighted mean using the set of samples S and the weighting method Mt q Quarter or index of any other division of the year S S 0 0 0 set of m strata ga Set of all samples S gsmpled j gUmsampled m agde Set of sampled strata 15 Suri Set of un sampled strata SOP OQ Computed weight of the catch in a stratum Q Sum Of Products St Stock a sub set of a species SW m S4 Set of weights associated with the set of sampled strata selected by method Mt SW S1 W me 29 V ye 2 v 28 W Q a Mean body weight in Catch of age group a in stratum Q w Mean Q The mean weight of all fish in the catch in stratum y Year Yield Q Yield Catch in weight in stratum Q Working group data Q Stratum Q St Ct Fl Ar y q combination of Stock St Country Ct Fleet FI Area Ar year y and quarter q QS Sampled stratum QY Un sampled stratum vVm Q a Weighting factor for weighted mean of age group a using method Mt 10 TABLES Table 1 List of options for weighting the calculation of mean proportions at age and mean weight in the catch Mt Method Vm 9 a Comment 1 Equal weight 1 0 No weighting 2 Catch Catch Q a Numbers caught Cannot be used if own sample is used 3 Landings LandingsN Q a Numbers landed 4 Discards DiscardsN
17. International Council for the Exploration of the Sea ICES CM 2001 P 23 Conseil International pour l Exploration de la Mer VPABASE A PROTOTYPE OF DATABASE FOR STORAGE AND PROCESSING OF VPA INPUT DATA by Per Sparre Ole Folmer and Clara Ulrich Danish Institute for Fisheries Research Charlottenlund Castle DK 2920 Charlottenlund Denmark email pjs dfu min dk ofo dfu min dk and clu dfu min dk Abstract The paper describes a prototype of database for storage and processing of Working Group input for fish stock assessment Input data are total catches by fleet and country combined with age distribution samples when available The processing of data applies the available age distribution samples to convert the national catches landings and discards into age compositions and sum the national data to total stock age composition which is the input to the VPA The database will maintain records on the processing of data as well as records on misallocated and unreported catches explaining differences between official data and so called WG data The database is implemented in ACCES EXCEL using a user friendly interface including a user s manual Data may be entered through userforms or transferred from other databases through a file with a specified exchange format The database was developed as a part of the EU funded project EMAS Evaluation of MArket Sampling strategies for a number of commercially exploited stocks in the North Sea
18. and development of procedures for consistent data storage and retrieval CFP Study Project 98 075 List of contents 1 Introduction eee ee ee e er be ss Ra a he aaa e e ceperit eh ee RO 1 2 Inputs to stock assessment 6 uie te dite it eser Pete Pen ERE ede reves PER te UE ER RR an 2 PEG LE cM 3 22 Weight At age conte eet I od e dee t tete er EO E iore Aloo 3 3 Estimation of catch and weight at age eene nennen nennen 4 3 1 lt Biological sampling ie ertt ee tero t e ree Fu e ee eee Veale 4 3 2 Sampled vs unsampled strata esses eene nnne nennen nnne 4 3 3 Estimating catch at age in unsampled strata sssssssssesseseeeeeeeeeneen enne 5 IAs SOP COMECH OD se 3 vire o d etie et ne otov ie a re heredes cale Corsa te ne er sh fe Lach 6 4 National data and ICES Working group data essere ener 6 4 Official statistics and working group data sse eee eren 6 4 2 Discards and landings eot e e ERE be eee a ei 7 4 3 Misreported and unallocated catch eene 7 5 Prototype Software apenas aidera rd tte d e be EEUU ETHER b eet 8 Del Tables andaelationg oett etetteteeen tipp tereti telues 9 5 275 USer IhteblaCe satietas RE ett M nce enna phaea feriae cu Rie eR MILI UA 9 SES Matitial data LASER 10 5 4 Automate data Input o sett re eet ete e ret alt a Pee he ree AEE ede e REUS 11 5 5 Processmg O dA iude ce er ester re o ebd ire e tria E
19. arge scale databases it was considered a suitable tool for the present application It allows having a user friendly interface Data may be entered through user forms or transferred from other databases through a file with a specified exchange format Reports from the database are made in EXCEL sheets A user s manual is available A number of software tools were already developed for data storage and combination of national catch information to the international level a In the ICES Herring Assessment Working Group for the Area South of 62 N HAWG national data are submitted on agreed spreadsheet forms The software called SALLOC is used to raise national data to the international level Patterson 1998 This Fortran program provides a log listing the actual fill ins used to cover all landings with missing biological information but this list is not given in the WG report Also no list is available in the WG report giving the extrapolations made of each national data set before submitting the data to ICES The Working Group provides a well documented overview of the sampling coverage from each country in the report b The Working Group on the Assessment of Demersal Stocks in the North Sea and Skagerrak WGNSSK keeps a database system to store and aggregate data on the flatfish species in the North Sea The database system FISHBASE consists of a user interface programmed in Pascal and a specific datafile format Data are stored in separate d
20. ata entry The first item in the main menu opens a data input form Figure 6 where data can be filled in manually One country record might contain several fleet records and one fleet record might contain several sampling records The form in Figure 6 is actually composed of three forms which correspond to levels of the hierarchy of tables see Figure 6 or Table 2 First level Annual catch by country The top part of the data entry form is related to the country catch record and the record selector in the bottom of the form tells the user which country catch record is current The record selector is the field called Record and the arrow buttons gives the user a way to go the first record move one step back move one step forward to the last record and to a new record Second level Catch by fleet and time period The second part of the data entry form contains the fleet record with fleet related information The fleet information in question relates to the country record in which the fleet record is embedded Ideally the sum of the fleet records over time periods of a country should sum up to the value of the country record Third level Age composition sample The third part of the form contains the age composition data samples relating to the fleet record in which it is embedded The sample is automatically given a unique sample number This number cannot be changed by the user and is only used to identify a sample for the
21. atafiles by country gear area sex if available and year Different levels of aggregations may be used in the datafiles ranging from files which have catch numbers weights at age and lengths at age by quarter sex gear area and country to files which only have total landings by country and year Raising is performed from the lowest level aggregation upwards The strong points of Fishbase are that it forces the user to correctly specify the input data and that data can be exchanged in separate datafiles so that national institutes can submit their data in those files Also it allows a historical access to the underlying data and other arrangements may be implemented to raise the data Negative points are that no manual is available that the program does not allow combination of areas that the input format is very difficult to generate with any other software than Fishbase itself and that there is no specification of reallocation rules c A series of basic programs are used to collect national age compositions and weights at age for roundfish stocks in the North Sea and west of Scotland The programs were developed in the 1980s in the Marine Laboratory Aberdeen and are operated from the dos prompt by invoking QBASIC AII data is stored in a common format An exchange format is used for exchange of data Strong points of the Aberdeen programs are that a rigid file structure is defined and that the raising procedure is carried through to gener
22. ating the input files for stock assessment Negative points are that there is a whole suite of programs which should be used in a specific order and that there are no explicit reallocation rules In the software presented here the raising of catch at age distributions to non sampled data is done automatically with the same algorithm as in Patterson 1998 But improvements of the current work are threefold Firstly this algorithm is combined directly with the database simplifying the manipulation and processing of data Secondly it is aimed to be a more generic and flexible tool that could be used for a number of different fisheries Thirdly it includes a higher level of data disaggregation to the fleet level This disaggregation is useful for improving the estimation of total catches at age by taking into account the differences of exploitation pattern arising from differences in fishing practices The database is thus intended to output the inputs to the VPA We consider here only the usual single species VPA used in most ICES Working Groups Multi species VPA although used for the assessment of some specific stocks is not considered at this stage of the database elaboration We do not consider the VPA tuning process which involves further specifications of time series of effort and catches for some commercial and surveys tuning fleets However it is worth noting that as the database is implemented at the fleet level so that future improveme
23. e catch SOP Sum Of Products is ama St SOR 2 Catch Q a We 25 a We 435 Catch 2 e 6 Catch C a A yin St 3 3 Estimating catch at age in unsampled strata The age composition within an unsampled stratum can be estimated using the age distribution of other strata which can be assumed to have a comparable pattern Let 27 be an un sampled stratum and let ON be a sampled stratum If we assume that weight at age and relative catch at age are the same for 27 and ON then we can estimate the total number caught in the unsampled stratum as Catch 27 e Yield 27 WE 27 7 and Catch 27 a E Catch 27 Y P aci 27 s a 8 It may also be preferable sometimes to use a subset of various sampled strata to estimate catch distribution in an unsampled stratum instead of only one as described in the previous equations The estimated distribution is thus a weighted mean of the age distributions in sampled strata The weighting coefficients might be uniform equal weight for all strata or linked to the importance of each stratum total yield number of fish measured aged inverse of variance etc The subset of sampled strata used for the estimation is denoted S 6 graa and has k elements S 103 05 O2 we define SW S a M OP a Vue 2 a V Mt Q a where SW is a set of weights vm Q a associated with the set of sampled strata and age group a selected by a weighting met
24. e stock caught by all fleets exploiting the stock Cvpa St y q a We use here the quarter as the time step by default as it is commonly used in the ICES area although assessments are afterwards conducted on an annual basis Any other division of the year for which data are available could be used The age index is comprised in the range amin Sf amax St where ay Sf 70 is the youngest age group considered in the VPA and amax Sf is the oldest age group considered in the VPA The oldest age group may or may not be a plus group The catch is taken by a number of fleets in a number of sub areas of the total stock area so that Ctyg Fly Ct Any Ctmax Flmax Ct AtMax Cypa Styqa gt gt M M Catch St y q a Ct F Ar M M Catch a 1 Ct 1 Fl l Ar 1 Ct 1 Fl 1 Ar 1 where Catch Q a Catch Landings discards in numbers of age group a in a stratum Q Fl 1 2 Flyax CO Ar 1 2 ATMax where Flyax Ct Number of fleets in country Ct and Aryax Number of areas The area can be ICES sub division ICES division or ICES area In a later version of the database the area concept could be generalized to include ICES rectangles and or specific areas defined by working groups Fleets are given homogeneous subsets of boats whose definition is made externally by the Working Group e g by fishing gear and or boat length class In general the dot e is used to indicate summation over one index e g when data are n
25. e to take its input directly from these tables The present proto type version of VPABase actually contains very few table Most of the tables listed here cannot be produced by VPABase as the system is now The authors of the present paper await the reaction and the wishes from the possible costumers of VPABase before they will start the work of designing output tables So what follows is what the present authors expect will become the requirements from ICES WGs but this still has to be tested by consultations The Excel reports of VPABase can be grouped as 1 Input data as provided by WG members and governments a Country records before processing b Fleet records before processing 2 Processing Tables showing the methods and input 3 Results of processing a Country records after processing b Fleet records after processing 4 VPA input Landings Discards and Catch In principle the database can display the data with any sorting or any grouping on the user choice Some tables however are expected to be in demand and for these there will be pre prepared routines to produce them These reports will comprise the traditional tables for official statistics of ICES working group reports as well as the non official data The example in Table 6 shows a table for all country records of a given 12 ICES area The system will be able to rearrange the tables on the users choice for example to give a report by country
26. g information on the data processing whereas the other tables contain observations from fishery ime for estimation T Age Distribution T Heet Catch T Country Catch _Use_OF Samples umber of Lengths Measureme sed Number In Age Gr ean Weight at age ey ICES Division ey ICES SubDivision Official Landings nreported Landings isMlocaed Landings Discards Sampled Catch lendkh measurements 22 T_County_Catch L Country T_Aeet_Catch Kev ICES Division Key ICES SubDivison Officia Landings Urreported Landings MisAllocated Landings ame OF Species Date of Recording eight of catch umber of Samples eighing method Ctoliths ength measurements 1 L_ICES Area k ey ICES Area x L_ICES_Division ke ICES Division ey ICES Area ame Of Division Mame Cf Region L_Quarter Key_Quarter L_ICES_SubDivision Quarter_Name ey ICES SubDivision ey ICES Division ame of SubDivision lias cf Subdivision L Stock Of Species Figure 4 Data tables Look Up tables and relations of the ACCESS implementation of VPABase omitting here the two auxiliary tales T Use Of Samples and T Use Of Samples Saved 23 1 T Age Distribution umber of Lengths Measureme Raised Number In Age Gr ean Weight at age L_Months ey Month ame L_Fleet_Of_Courtry L_Use_Of Catch i ame Of Use Of catch
27. he list includes characteristics of ALKs Age length keys such as number of length measurements and number of ALKs 3 4 80 P correction The SOP is the sum of the products of catch numbers at age times the weights at age a Max St SOP caren 2 X Catch Q a wec Q a 11 a a min St Ideally the SOP should be equal to the Yield 2 which was given as input If that is not the case the equality can be achieved by applying a SOP correction either to the catch or to the weight a Max St Yield Q gt Catch Q a Correction 2 Wo 2 a a a min St 12 where Correction 2 melden SOP Catch 42 The SOP correction was discussed by Lewy amp Lassen 1997 who concluded that the SOP is an unbiased estimate of the total catch and that SOP corrections should not be made We consider four options in the case of SOP discrepancies 1 Correct the catch at age 2 Correct the weight at age 3 Correct the input Yield replace the value of input Yield by the value of SOP 4 Do nothing According to Lewy and Lassen option 3 is recommended In the following it is assumed that any SOP problem has been resolved before the data are entered into the database If however there is a discrepancy the value Yield will be automatically changed to the SOP Option 3 4 NATIONAL DATA AND ICES WORKING GROUP DATA The annual yield by country summed over fleets and ICES area are reported by member governments to
28. hod Mt The weighted mean proportion at age in the unsampled strata 1s then DA Pres 25 a id Vat Q5 a Presa 27 8 S Mt 9 Y Preaa 2 2 v 05 2 a a min St OS The calculation of a weighted mean proportion at age is only possible when the weighting factors are available for all strata in S The methods Mt equal weight all v 1 0 or Mt yield are applicable for any set of strata The method Mt number of otoliths requires having an associated number of otolith measurements for all strata Note that the procedure to calculate weighted mean proportions at age may also be applied to sampled strata The otolith samples taken from a catch stratum are usually used to estimate the age composition of that stratum However if the sample is very small or for some other reason suspicious it may be better to use alternative otolith samples or to combine the suspicious sample with samples of other strata The weighted mean body weight in the catch is then 3 Wc 02 2 v 29 2 Ww Mean Q27 a S Mt Q5 es i i Y 2 2 QS es 10 Then we can calculate the total number of individuals caught using Eq 7 and then the number of individuals per age group using Eq 8 A list of possible weighting options for the calculations of proportions at age and mean weights in the catch is presented in Table 1 The list of options for weighting factors may be extended in future versions of VPABase Obvious extensions of t
29. ield St Fl Arq 16 FleFleets Ct AreICES Div q 1 This equation must be met by all input data of each country If it is not met the Yield St Fl Ar q will be corrected by the database system using the country and area specific correction factor Landings LC ex St Ct Ar CorrectionWG St Ct Ar oe 17 B Yield geforeCorrection St FI Ar q FleFleets Ct AreICES Div q 1 where Yield peoreCorrection 9t Fl Ar q and Yield after correction respectively St Fl Ar q stand for Yield before and AfierCorrection Yield St Fl Ar q Yield St Fl Ar q CorrectionWG St Ct Ar 18 AfierCorrection BeforeCorrection However basically working group members should supply the data so that no correction would be necessary The program will always compute the correction factor and the user should check the value The database will keep the original value before correction for the purpose of documentation S PROTOTYPE SOFTWARE A prototype database system was developed under the working title of VPABase The prototype version of VPABase was implemented in Microsoft ACCES and EXCEL version 2000 The term Prototype means a Working model in the early stage of the design process The prototype does run and executes some of the tasks specified in the foregoing sections The prototype however is far from the stage where it is safe to distribute to ICES WGs It has not yet been conscientiously
30. ime period combined with age distribution samples when available It was developed as a part of the EU funded project EMAS Evaluation of MArket Sampling strategies for a number of commercially exploited stocks in the North Sea and development of procedures for consistent data storage and retrieval CFP Study Project 98 075 Pastoors et al 2001 This project intended to evaluate the adequacy of the international market sampling effort for a limited number of commercially exploited fish stocks in the ICES area to quantify the uncertainties in estimated catch in numbers and mean weight at age to advise on appropriate sampling levels and methods and to define protocols and develop prototype software to store and aggregate national sampling data The data processing uses the available age distribution samples to convert the national catches landings and discards into age compositions and sums the national data to total stock age composition which are the main input to the VPA Virtual Population Analysis The processing of data implies filling in gaps in data caused by some countries not supplying full sets of input data The gaps are filled in by extrapolating samples data to non sampled catches data Figure 1 The database is a relational database implemented in ACCESS chosen for its simplicity of use and because it is a standard software commonly used in fisheries institutes Although ACCESS might not be designed for storing and manipulating l
31. is Un Discards Ho of Hoof Procent of Division landning Allocated reported fleets sampled catch fleets sampled Iva Iva East 3379 0 0 0 4 0 10 IVa Iva East 14154 3506 0 0 4 1 166 Iva Iva Wes 652 0 0 0 4 0 10095 Iva Iva Wes 22152 3506 0 0 4 0 116 Ivb All 3906 0 0 0 4 0 3096 Ivc All 10648 3506 0 0 4 0 13396 Country Netherlands Division Sub Official Mis Un Discards No of Hoof Procent of Division landning Allocated reported fleets sampled catch fleets sarrpled IVa Iva Wes 1000 100 100 200 1 1 1400 Table 5 Inventory of fleet records LANDINGS DISCARDS Total yield Number of Coeff of Total Number of Coeff of otoliths Vari ance discards otoliths Vari ance Fleet 1 Fleet 2 Fleet x Country B Fleet 1 Fleet 2 Fleet x Fleet 1 Fleet 2 Fleet x Table 6 Report proto type Country records Official and non official catch data Country records of ICES Division No X Official Landings Unreported Mis allocated Discards Country 1 Country 2 Country X 18 Table 7 Report proto type Fleet record sample Catch at age weight at age and additional information on sampling number of otoliths and coefficients of variation Country A Fleet a ICES Division No x sample No y LANDINGS DISCARDS Age gr TOTAL Number Body wgt Numberof Coefficient Number Body wgt Numberof Coefficien
32. le is selected for a given catch The default method is to give equal weight to each sample but if additional information i e numbers of otoliths is present this can be used when the samples are combined After the selection has been made the user moves back to the processing menu where she he can save the selection to a permanent table data table T Use Of Samples Saved see Figure 3 so the selection scheme is stored for later documentation of the process Default for the creation of the table is that if a catch record has a sample this sample will be used for this catch but there is the option not to use the default Option 2 above 5 6 Reports in Access format The ACCESS system can create reports or output tables on its own The ACCESS format tables however are not very flexible with respect of additional processing and change of style see example in Table 4 Tables in EXCEL format is believed to be more attractive for members of ICES working groups Reports in Access format are inventory reports of the database They contain information on the number of records and other features of the input data and the national sampling programs Prototypes of the inventory reports are given in Table 4 country records and Table 5 fleet records 5 7 Reports in Excel format These are the standard tables suggested for potential inclusion in the ICES Assessment working group reports Ideally the assessment software should be abl
33. not Name in ACCESS T Use of Samles see Fig 3 Name of Field Comment Key Catch Age Pointer to age composition record sample Key Catch Pointer to fleet record Yes No Yes Ifa sample is selected to represent a catch that is if the record Sample selected Key Catch Age is selected to represent the fleet record Key Catch Yes No No If no sample age composition record has been selected for the fleet INo sample Record Key Catch ACTUAL USE OF SAMPLES FOR ESTIMATION OF AGE COMPOSITION Data table containing information on the actual selection of samples to represent fleet records catches Contains only combinations of fleet record and Age composition Record which are selected for estimation of age composition Name in ACCESS T Use of Samles Saved see Fig 3 IN ame of Field Comment Key Catch Age Pointer to age composition record sample Key Catch Pointer to fleet record Yes No Yes Ifa sample is selected to represent a catch that is if the record Sample selected Key Catch Age is selected to represent the fleet record Key Catch Weighting method Look up Method used to compute weighted average of two or more samples Time for selection The time and date the selection was made in the database 17 Table 4 Inventory of Country records Example of report in ACCESS format Inventory of Country Records Country Denmark Division Sub Official M
34. nts could easily be provided to store tuning data as well in the database In this paper we firstly summarise briefly the main inputs to be used in the VPA and how they can be calculated from sampling data We then summarise how and why these inputs are modified during working groups before being used in assessment This aims to present the full set of parameters that should be included in the database The database itself is presented afterwards 2 INPUTS TO STOCK ASSESSMENT The VPA compute stock numbers and fishing mortality coefficients by age and year from data of total catches in number and estimates of natural mortality coefficients by age and year although natural mortality is most often considered as constant over years Biomass is thus calculated by multiplying stocks numbers by weight at age To facilitate the notation we introduce the Stratum By a stratum is meant a combination of a stock St a country Ct a fleet F an area Ar and a quarter q in a year y We use the symbol Q for a stratum Q St y Ct Fl Ar q The symbol S represent a set of strata S Q do CR T where m is the number of elements in the set S Appendix 1 contains a complete list of symbols used in the formal description 2 1 Catch at age The input to single species VPA without tuning are the numbers caught landed and discarded of stock St by age group a and during quarter q in year y from the total area of distribution of th
35. of VPABase input data may also be transferred in ACCESS table formats from one replicate of VPABase to the master version of VPABase which operation should be managed by an international data coordinator 5 5 Processing of data The third item in the main menu Select samples from un sampled catch and process data opens a submenu Figure 7 for data processing If a fleet record has an age composition sample attached to it is a sampled catch the user does not need to take further action as far as processing is concerned In that case the VPABase will make the estimation of the age composition of the total catch from that sample The user need only take action in the two cases 1 When no age composition samples is available 2 When the user wants to use additional age composition samples for a sampled catch The action to take 1s to select one or more samples to represent the unsampled catch If more than one sample is selected the user will also have to specify which method should be used to pool the selected samples how the weighted average should be calculated Before any actual selection of samples can be made the user has to create a table to hold the tables of potential selections the data table T Use Of Samples See Figure 3 After having created the table of potetial samples the user can move to the next item on the processing menu and start making selection for un sampled strata Item 2 Select samples
36. ose a n Eee ree en eee eet 11 5 6 Reports in Access TOEImat uie eee etr sau EEEE Te rer AERTS EENE haee vae T 12 5 7 cReportsan Excel formi t a ue cote Db rte fex RU e DESC ous vanes eaaet ies 12 6 Data storage and data exchange cccccesscesseessecessceeeeeeeceeseeeseeeseeescecsaecsaecsaeceaecesecnseceaecnseenseeeseenes 13 7 Discussion and conclustofis deerit etr ee otn REFERT ORO Uere pO Vite Raves 14 8 hom cte EE 15 9 LIST OE Syrabols ecrire de ts octet rele ite rede tete D E Eee E E 15 10 Tables esac ine S rm te EO dro eec eee dre rte rue 16 BEES OIL NE ME 21 List of Figures Figure 1 Illustration of the gap filling process sss enne eren eene 21 Figure 2 The basic hierarchy of tables in VPABase sse enne enne enne 21 Figure 3 Data Tables and relations of the ACCESS implementation of VPABase sess 22 Figure 4 Data tables Look Up tables and relations of the ACCESS implementation of VPABase 23 Figure 5 Main menu of VPABase en aa a eene enne erret in rennen innen nene enne 24 Figure 6 Data entry form of VPABase ssssssssesesseeseeeeeee enne entente entren irren enne rennen nnns 25 Figure 7 Menu for data processing ccssccssscesscessceeceeceseceeeeessecesecseeeeeeeseaeceseeeseecseecsaeesaecaecesecnaeenseenes 26 Figure 8 Data entry form for the selection of age compositions to be used to fill in unsampled catches 26 Figure 9 EXCEL worksheet used
37. ot disaggregated by quarter but are available only on an annual basis 4 Catch St y e a Ct Fl Ar X Catch St y q a Ct Fl Ar q 1 2 2 Weight at age Estimated stock numbers are converted into stock and spawning stock biomasses by using weight and maturity coefficients by age and year Predictions of yield are made from mean fishing and natural mortality at age and mean weight at age The weight at age is thus a major parameter of the assessment procedure It relates to a specific time e g the beginning middle or mean over the time period In subsequent equations the weight at age 1s noted W St y q a Ct Fl Ar W Q a Mean body weight in Catch of age group a in a stratum Number of fish caught is calculated as Yield 2 a Catch Q a We Q a 2 Where Yield Q a is the catch in weight of stock St and age a in a stratum The total yield Yield Q must be available for all strata to make the following estimation feasible 3 ESTIMATION OF CATCH AND WEIGHT AT AGE 3 1 Biological sampling Catch and weight at age are estimated from biological sampling data which provide detailed information on length and or age composition of the landings in a stratum Data include the number of otoliths read and the number of fish measured by age The associated coefficients of variation may be available Sampling data by age and stratum in the database are thus na Q a Number of age measurements otoliths
38. ot filled in if month filled in Look up Sub set of fishing harvesting sector of country Look up Subset of Species Look up Discards Landings The date the record was entered or modified in the database Weight of landings or discards catch landings discards is not input INumber of samples Number of age composition samples 0 if no samples available Index for weighing method Default value 1 for equal weight for each sample Method of weighing pdated when new selection is made Otoliths Sum of number of otoliths updated when new selection is made length measurements Sum of numbers of length measurements updated when new selection is made AGE COMPOSITION RECORD SAMPLE Name in ACCESS T Age Distribution see Fig 3 Name of Field Comment ey Age Distribution Unique key of age sample record ey Catch Pointer to fleet record one to many ge group 0 1 oefficient of variation Variation of estimate of this age group e g From bootstrapping umber of otoliths Number of otoliths in age composition sample in this age group otal number in catch In discards or landings raised numbers ean weight Mean weight of age group during time period in question POTENTIAL USE OF SAMPLES FOR ESTIMATION OF AGE COMPOSITION Auxiliary data table to contain information on the potential selection of samples to represent fleet records catches Contains all combinations of fleet record and Age composition Record whether selected or
39. pled strata Table 7 The table will contain information on the processing extracted from Table 9 The results for the country is either the sum over fleets or the estimation based on borrowed catch distributions Table 10 The input to the stock assessment process is the summation over country records Table 11 The only EXCEL report produced by the proto type version of VPABase is the VPA input as shown in Figure 10 6 DATA STORAGE AND DATA EXCHANGE Data are stored in tables and relations in the relational database as described above It is tradition in ICES to store data in so called lat files that is a data storage where each data record can be read independently and where no relations between data are specified Data stored as flat files usually takes up much more space than data stored as tables with relations Should a flat file be required for example for analysis by the SAS the database can easily produce such a flat file Data can be submitted from countries to the central database in three ways 1 Entry of data in replicate database 2 Entry of data in EXCEL format Using a fixed standard set up 3 Entry of data from screen through the data entry forms The data that is presented in Excel format will be imported by the data base and added to the data already on the system Alternatively the database could have been implemented on the internet and data could be entered directly in the central database This solution i
40. positions to be used to fill in unsampled catches hypothetical example amp Select age distributions p ial xl SELECT SAMPLED CATCH RECORDS FOR VPABaSe ESTIMATION OF AGE COMPOSITION OF Catch record Exit Selection form ils Country Denmark ICES Subdivision IVa East 1 T IV Species Herring ICES Division Iva C Weighted by number of otoliths Year 2000 ICES Area Iv Quarter 2 Official Landings 3379 Weighted by sampled catch Fleet Industrial Fishery Weight of catch 0 C Weighted by sample size i p C Weighted by inverse CV age Use of catch Reduction NO age distribution g y age Catch no 107 for this catch C Weighted by inverse CV weight Catch records with samples Number of measurements No Country Area Div Sub Div Qrt Species Fleet Use of catch Samples Lenght Otoliths select 126 Denmark IV IVa IVaEast 1 Herring Humancons H C 1 a O 130 Denmark IV IVa IVaWes 1 Herring Humancons H C 0 0 0 O 133 Denmark Iv IVa IVaWes 4 Herring Humancons H C 0 0 0 O 135 Norway IV Iva IWaEast 2 Herring amp Reduction 25 0 0 E 136 Norway IV IVa IWaEast 3 Herring A Reduction 6 0 0 E 137 Norway IV IVa IvaEast 4 Herring Reduction 0 0 0 O 138 United State IV Wa IVaWes 1 Herring Industrial fit Reduction 6 329 59 Iv 139 United State IV IVa IVaWes 2 Herring Industrialfis Reduction 3 6788 237 Iv 140 United State IV IVa IvaWes 3 Herring A Reduction 0 0 0 O 141 United State IVa Wes 4 Herring Reduction 1 O
41. s rather straight forward to implement The supply of data in a traditional flat file format e g SAS files will be accommodated in the future The proto types of the Excel data exchange formats corresponds to the tables of VPABase Country file Table 12 and Fleet age file Table 13 The fleet table and the age distribution table of VPABase have been combined into one exchange file The fields in this table may not all be filled in For example if catch at age and weight at age are given the landings or discards are given as the sum of products and these cells do not have to be filled in If they are the SOP will have priority over the landings discards in tonnes This version of VPABase is not an online system and when a record is updated it is not immediately updated in the international network Each institute of participating in the ICES assessment working 13 group will get a replica of the VPABase A replica is simply a copy of VPABase for which the design cannot be modified Before the annual WG meeting the master VPABase will be synchronised with all the replicas To synchronise two replicas of a database means to exchange all new records and updates of records This task could be carried out by a data coordinator 7 DISCUSSION AND CONCLUSIONS A general conclusion from the whole project EMAS Pastoors et al 2001 was that the sampling procedures in the different countries have very different characteri
42. stics adapted to the local circumstances the availability of sampling options and the behaviour of the fishery There is no obvious method for standardization of sampling procedures available The implications of this uncertainty on stock assessment results are relatively minor when considering only the total catch at age matrix but are larger when using commercial CPUE to tune the assessment model The storage of all basic data used in the estimation of catch at age 1s of great importance to keep track of the procedures used and record the consistency of sampling data This is not yet the case in ICES where different working groups use different methods and software and where catch data are exchanged only at the national level of aggregation and with different format The database presented here should therefore improve the homogeneity standardisation centralisation transparency and flexibility of data used in assessment It is generic and thus suitable for a great variety of fisheries It allows accounting for various levels of aggregation and various scales of basic data It gathers both the database and the program performing all calculations until the final VPA input within one single software with a great level of specification provided by the user It has been designed with a great user friendliness and easiness to use It is not a rigid structure as all user can have access to the programming code and possibly improve it for his own use
43. t Number landed Landings otoliths of variation Discards Discards otoliths of variation caught Lo J o op op o p 2 S 1 Max a TOTAL SOP Mean Wgt Total SOP Mean Wgt Total If data above are not available then total catc landings or discards in weight may be available Not avail Not avail il Not avail Fleet record No X Fleet record No Y Method no Method no Sam ICES Land Disc Landings Discards Landings Discard ple Ct Fleet Div Samp Samp Samp Used Sampl Used sed Sampl Used mE Fleet a ICES Division No x Number Body wet Number Body wet Age gr landed Landings Discards Discards 0 1 Max a TOTAL SOP Mean Wet SOP Mean Wet Method no Landings Samp Discards Sampl Used Used SOP Sample 19 Table 10 Report proto type Country total of fleet records Country A All Fleets All ICES Divisions LANDINGS DISCARDS TOTAL Age gr 0 Number Body wgt Numberof Number Bodywgt Number of Number landed Landings otoliths landed Discards otoliths caught 1 TOTAL SOP Mean Wet SOP Mean Wet Total SOP Table 11 Input table to stock assessment last column Table 12 Proto type of Country record for data exchange COUNTRY RECORD IV Given only if Div or sub div not filled in Division IVc Given only if sub div not filled in Sub division IVc East 2001 Table 13
44. tested and so far no feedback has been received from the potential users the ICES WGs Thus an unknown number of man months of development testing and consultations with the clients are still pending before public distribution should be made A high degree of user friendliness is aimed at in the design of user interface and the output options of VPABase This feature has been attempted from the early stage of the development Primary versions of a user s manual and a reference manual of the VPABase has been prepared This Section summarizes the two manuals The database can read input data from Excel formats and also produce reports in Excel format The example of EXCEL work sheet used in the present version of VPABase has been adopted from the spreadsheets used by the Herring WG and the Mackerel Horse Mackerel Sardine and Anchovy WG ICES 2000 Input and output can also be handled out with plain ASCII text files comma separated The VPAbase is constructed as a relational database according to the standard database design rules normalization rules 5 1 Tables and relations The tables and the relations of VPABase consist of Input data tables Look up tables and Auxiliary tables The look up tables are used to select standard input values whenever possible as that will reduce the risk of mistakes on data entry The principal data tables of VPABase contains the numbers at age and weight at age distributions and
45. time The best solution would be to provide an Internet access to VPABase allowing using it from any kind of remote computer Finally the database should be extensively tested by a number of working groups and fisheries research institutes before being validated as a standard tool So far the database has been tested only at DIFRES Danish Institute for Fisheries Research We believe that this tool could be of great help within ICES working groups and we hope that it will encounter interest enough to be tested and used 14 8 REFERENCES ICES 2000 a Report of the working group on the assessment of mackerel horse mackerel sardine and anchovy ICES C M 2000 ACFM 5 ICES 2000 b Report of the Herring Assessment Working Group for the Area South of 60 N ICES CM 2000 ACFM 10 Lewy P Lassen H 1997 Should total landings be used to correct estimated catch in numbers or mean weight at age ICES J Mar Sci 54 462 470 Pastoors M A O Brien C M Flatman S Darby C D Maxwell D Simmonds E J Degel H Vinther M Sparre P Vanderperren E 2001 Evaluation of MArket Sampling strategies for a number of commercially exploited stocks in the North Sea and development of procedures for consistent data storage and retrieval EMAS Final Report of the EU study no 98 075 RIVO The Netherlands CEFAS UK DFU Denmark SOAEFD Scotland CLO DZ Belgium Patterson K R 1998 A Programme For Calculating Total
46. with each row representing an ICES area Table 7 shows a proto type of fleet record table There may or may not be age weight composition data available for a stratum If age weight composition data are not available then total catch or landings mat be available in which case only the last line in the table will be filled The fleet records have unique samples numbers Sample No y which are used for of identification of the fleet records when used for estimation of non sampled strata An important requirement for the database system is to document the age composition estimation procedure To document the process of using age compositions from other strata means to record from where the borrowed information originates and which method of weighting was applied Table 8 shows a report which for each fleet record contains a log of its processing Each row in the table represents a stratum The first column gives the unique index of the stratum The following 5 columns gives the particulars of the sampled stratum The columns Landing Samp and Discard Samp Indicates if these catches were covered by the sampling program The following columns contain the selection of borrowed samples used for estimation of landings and discards age compositions The first line contains the methods of weighting the data if average values of several strata were used For the processed fleet records the output will be a table similar to the output table for the sam
Download Pdf Manuals
Related Search