Home

SurvCalc manual 12March2012 - These are not the droids you are

1. b calc_biomass calculate biomass 1 calc_LFs calculate LFs B calc_biomass_and LFs calculate biomass and LFs 2 phase_2_calc do calculations for allocating phase 2 stations in a 2 phase survey o output_flat_files only output flat files of data to be used only in conjunction with arguments s t u or v p output_LW_coeffs output a table of length weight coefficients from database rdb c output_for_catch_at_age output a file in either survey or survey sub h help format whichever is appropriate for input to the catch at age software list all arguments L show license display the SurvCalc end user license In addition one or more of the following run time arguments can be used to provide information about input and output files Argument S infile T infile U infile V infile W infile X infile s outfile t outfile u outfile v outfile Description read stratum data from a flat file rather than a database default value for infile is stratum in read station data from a flat file rather than a database default value for infile is station in read catch data from a flat file rather than a database default value for infile is catch in read length data from a flat file rather than a database default value for infile is 1gth in read subcatch data from a flat file rather than a database default value for infile is subcatch in read combined station and catch
2. calc_biomass or calc_biomass_and_LFs They could also be used with task calc_LFs if the LF scaling was specified by adding a command like LF_scaling numbers_per_hour Example 1 The first example is for the simplest analysis for one species in one trip the trips tan0617 species tan0617 codes HOK database Empress where preferences distance_towed r vulnerability command in this file won t usually be needed but is included in all three examples because it illustrates a couple of useful points input_from_database t_station gear_perf lt 3 and station_no 60 corded_distance recorded_speed time from_lat_long width_swept recorded_doorspread catch_weight recorded calculated 27 vulnerability tan0617_HOK default_value 1 other_stations 71 72 other_values 0 8 0 8 sub_populations sexes all all Lmin 0 60 Lmax 59 200 labels to60 60 output_tables sub_biomass_by_stratum T biomass_by_species F biomass_by_species_stratum F LFs_by_stratum T LFs_by_station F Number_measured T LF_totals T output_precision quantity density biomass LF_number cv type dec_place dec_place sig_fig dec_place precision 0 0 8 0 Example 2 The second example extends the first to analyse multiple species in one trip It differs from the first example in only two places The first at subcommand codes of command species specifies the additional species to be analysed The second is r
3. SurvCalc User Manual v1 2 2011 09 28 R LC C Francis D Fu NIWA Technical Report 134 ISSN 1174 2631 March 2012 Published by NIWA Wellington 2012 Enquiries to Science Communication NIWA Private Bag 14901 Wellington New Zealand ISSN 1174 2631 NIWA 2012 Citation Francis R I C C Fu D 2012 SurvCalc User Manual v1 2 2011 09 28 NIWA Technical Report 134 54 p The National Institute of Water and Atmospheric Research is New Zealand s leading provider of atmospheric marine and freshwater science Visit NIWA s website at http www niwa co nz Table of Contents 1 si INTRODUCTION isscsses cess stitess testes nccteeeniatens oc steessaesbeunnn caetee sn ai iiini iiei 5 1 1 OVERVIEW senei ates E E E E E E E E 5 1 2 Relationship of SurvCalc to trawlSurvey ooooconccnnccnocnnocnnnoncnoncnn nono nononocnno conc cnnccnnnannos 6 1 2 1 Extensions to trawlSuUrvey scescesecsseceseceeeceseeeeeeeseeseneeeaeecaeecaaecaaecaesnaeenaeens 7 1 2 2 Excluded features of trawlsurvey ooooconocococococcnonnconnnonncnnncnnn nono ccoo conocio conan nccnna ns 7 1 2 3 Corrections to trawlSUrVey sceeeeeeseseceseceeseeeseecseecscecsaecaecseesseceseeesneeeneeees 7 1 3 Possible future extensions to SurvCal 0 0 ee eeceseceeseeeseeeneeceaeceseceaeceseeeeeeeseeseneeees 8 RUNNING Surv Calcuta A A bras ie tna 9 INPUT FILE SPECIFICATIONS 05 ccscecdssesssetsctesencatedueseutebenadshessiees
4. for every combination of trip and species in the input file a file that can be read by the catch at age function import length data as well as a main output file myout in the above example whose format is described in Section 4 1 The catch at age files are either in the survey or survey sub format with the latter format being used if there are any subcatches for the given trip and species The names of the catch at age files are not specified by the user They are constructed from the species and trip names in the form LFttttttccc i for format survey or LFttttttecesub i for format survey sub where tttttt is the trip code ccc is the species code and i is an index used by SurvCalc to ensure that any existing files don t get overwritten For example the first time you do a SurvCalc c run for HAK in trip tan0901 SurvCalc will produce the output file LFtan0901HOK 1 or LFtan0901HOKsub 1 if there were any subcatches If you then do another SurvCalc c run for the same species and trip in the same directory without deleting the first output file then the new output file will be LFtan0901HOK 2 or LFtan0901 HOKsub 2 The survey format is as follows comment line trip number_of_strata length weight parameters male_a male_b female_a female _b optionally unsexed_a unsexed_b one row for each stratum stratum_name stratum_area n_tows and then a table of length records with a
5. be repeated in the main input file are discussed in Section 3 1 7 3 1 1 Commands defining the data The commands in this section define what type of survey is to be analysed trawl or pot which trip s and species are to be analysed whether the data are to be read from a database or flat files and if the former which stations to use from the specified trip s If the data are to be read from a database some of all of the data in Table 1 will be extracted Which tables are extracted depends on the specific analysis Table 1 Database tables from which data may be extracted by SurvCalc and the attributes extracted from each table The database and table names given are those for the original implementation on Empress at Greta Point slightly different names might apply in other situations Database Table Attributes extracted trawl t_stratum trip_code stratum area_km2 trawl t_station trip_code station_no stratum distance lat_s NorS_s long_s EorW_s lat_s NorS_f long _f EorW_f speed dist_doors and any other attributes included in input_from_database t_station_columns trawl t_catch trip_code station_no species weight trawl t_subcatch trip_code station_no species subcatch_no weight trawl tleth trip_code station_no species subcatch_no lgth percent_samp no_a no_m no_f rdb lw_coeff spp_code sex lw_coeff_a lw_coeff_b lw_coeff_c This table extracted only for task output_for_catch_at_age see Sect
6. discussed in Section 4 4 4 1 Main output file The structure of this file is most simply illustrated by the following which shows the structure of the main output file expected when the input file is as in the last example in Section 3 1 6 and the task is calc_biomass calc_biomass_and_LFs or calc_LFs the indenting in what follows is simply used to illustrate the structure of the file it does not occur in the actual file Header SELECTIONS for trip tan0617 PREFERENCES for trip tan061 7 Results for species HOK in trip tan0617 Results for species HAK in trip tan0617 Results for species LIN in trip tan0617 Tables for trip tan0617 SELECTIONS for trip tan0714 PREFERENCES for trip tan0714 Results for species HOK in trip tan0714 Results for species HAK in trip tan0714 Tables for trip tan0714 For tasks output_flat_files and output_for_catch_at_age the main output file contains only the header For task output_LW_coeffs it contains the header plus a table of length weight coefficients The header contains general information about the particular run of SurvCalc including the command line the date the version of SurvCalc this is useful if bugs are subsequently found in the program the user name and if the survey data have been extracted from a database the SQLs that were used to do this If LFs are calculated then the last line in the header describes the LF scaling The SELECTIONS section contains two lists one of the selecte
7. equation 21 equation 18 from lw_coeff Table 4 continued j indexes the records in table t_lgth for station i and the specified trip and species so the following variables are derived from attributes in this table Variable Definition How calculated where from Pij proportion sampled percent_samp 100 Nijx number of fish by sex no_m k 1 no_f k 2 no_a no_m no_f k 3 no_a k 4 Ly length of fish cm lgth Win weight of fish g Waal l indexes fish length Wr weight of fish of length and sex k g W ate ni SN refers to scaled numbers of fish MSN is the mean scaled numbers SN is scaled numbers corrected for consistency with biomass estimates and length weight relationships SN of length and sex k at station i equation 13 SN ay MSN six SN s of length and sex k in stratum s equations 14 or 16 17 and 20 SNy SN s of length and sex k equation 22 23 TN refers to total numbers i e LF totals which are sums over length of scaled numbers TN x for station i TN gt SN TN for stratum s TN SN amp equation 28 DS i LF TN MTN x for stratum s MTN gt MSN m m TN for the whole population TN 4 Ne SN i Table 5 Conventions used in summations See Section 5 7 for a further convention used only for task output_for_catch_at_age Symbol Range of summation the n stations in stratum s the m stations in stratum s in which lengths were measured E 5 stations in which eithe
8. header row as follows stratum tow kg km2 length no_m no_f no_t and a whole bunch of data rows The survey sub format is the same as this except that the final table of length records has header row stratum tow subsample kg km2 subsample_kg km2 length no_m no_f no_t and two additional columns subsample subcatch number and subsample_kg km2 37 4 4 Precision of numbers in output files The degree of precision expressed as a number of significant figures or decimal places for each of five types of output quantities is controlled by command output_precision Table 3 The default levels for each output type were set on the basis that c v s are unlikely to fall below 1 and are usually much higher Note that negative numbers can be used for decimal places e g 12345 67 presented to 2 decimal places is 12300 Table 3 Description of the quantities whose precision is controlled by command output_precision and the default level of precision a number of either significant figures or decimal places assigned to each Default Quantity precision density 4 sig figs biomass 4 sig figs LF_number 4 sig figs cv 1 dec place gain 0 dec place Description all columns of type XYZ_kg XYZ_kg km or XYZ_kg km2 in station data and stratum data output files column kg km2 in stratum_summary table column kg km2 in catch at age output all biomasses including lower and upper bounds i
9. in flat files or in Empress or Postgresql databases A third more subtle difference between the two programs is that the format of the output has been tweaked so that it is easier to read it into R for plotting and further analysis see Appendix 2 All the main calculations in trawlsurvey of biomass and LFs are exactly the same in SurvCalc in fact much trawlsurvey code was reused in SurvCalc However SurvCalc includes several new features see Section 1 2 1 discards a few features of trawlsurvey see Section 1 2 2 and corrects a couple of minor errors in that program see Section 1 2 3 Before describing these differences in functionality of the two programs it s worth noting for the record some technical programming differences trawlsurvey is actually the combination of two programs an Empress 4GL interface which is what the user sees this generates all the blue screens and a C program which is run in batch mode from that interface SurvCalc is a single C program 1 2 1 Extensions to trawlsurvey The following are the main features of SurvCalc that were not possible with trawlsurvey excluding those just mentioned Biomass and LFs can be calculated for multiple trips and or species in a single run Data can be extracted from the trawl database and output in flat files of station data one line per station stratum data one line per stratum catch data and length data Input files for use in
10. ncnnnccnnccnnacns 52 Appendix 2 The SurvCale R library ooonconnncnnncnoncconncnononononannnnnnnnn ccoo ccoo acon nono nccnacnnecns 53 1 INTRODUCTION 1 1 Overview SurvCalc is a C computer program which analyses data from stratified random surveys Its primary purpose is to calculate estimates of biomass and or length frequencies LFs and associated coefficients of variation c v s from survey data These data may be held either in a database structured like the Ministry of Fisheries database trawl Mackay 2000 or in flat files SurvCalc supersedes and uses some code from the program trawlsurvey Vignaux 1994 Users of SurvCalc are urged to include their input files in an appendix to any report describing the analysis of stratified random surveys The main input file for SurvCalc has been designed so that taken together with this manual 1t fully documents all the choices the user makes in calculating biomass etc e g the choice of stations to include and how distance towed is calculated if there is no recorded value This will allow readers of survey reports to replicate the analyses therein When SurvCalc is run using data from flat files rather than from a database these flat files should also be included in the report to complete the documentation Each time SurvCalc is run it carries out one of the seven following tasks The first three tasks involve different types of calculations that may be made either during
11. number of stations allowed in each stratum and mydat should be a dataframe containing the historical data in the format stratum survey catch 0001 tan0303 Sze 0001 tan0303 2D 0001 tan0303 12 1 0002 tan0303 4 2 0002 tan0303 2 48 0001 tan0402 bs where the last column contains catch rates e g t km or t km etc The output from allocate is a list with three components cv the estimated c v achieved by the calculated allocation stations the total number of stations and allocation the allocation of those stations to the strata For example the output from the first call above might look like the following Scv 1 19 8 Sstations 1 32 Sallocation stratum stations 1 0001 17 2 0002 12 3 0003 3 which would indicate that 32 stations with 17 12 and 3 in strata 0001 0002 and 0003 respectively could be expected to achieve a c v of 19 8 It is not uncommon for surveys to have separate target c v s for two or more species In this case we would need to construct a different dataframe of historical catch rates for each species and run allocate with each of these in turn In this case it is of interest to find out what c v would be expected for one species if we used the optimal allocation for the other species This is easily done by modifying the strata dataframe so that for each stratum the minimum and maximum numbers of stations are both set equal to the optimal number It s important to
12. produced for each species trip combination One line the first with label a11 for the total biomass and then one line for each requested sub population with user supplied labels Lower and upper bounds are estimate 2 s e s population lbound biomass ubound cv Nstations all 31601 46081 60561 16 97 to60 114 846 1807 57 90 60 11830 20001 28172 20 90 sub_biomass_by stratum Optional table produced for each species trip combination Contains biomass estimates and c v s for a single species by sub population and stratum the column labels are user supplied for each sub population stratum to60 cv_to60 60 cv_60 0001 726 66 639 85 0004 87 71 3345 50 biomass_by_species Optional table produced for every trip in which more than one species is analysed Like the biomass table above containing one line total biomass per species species lbound biomass ubound cv Nstations HOK 31601 46081 60561 16 97 HAK 1827 2628 3429 15 97 34 biomass_by_species_stratum Optional table produced for every trip in which more than one species is analysed One line per stratum containing biomass estimates and c v s for all species analysed for that trip stratum HOK cv_HOK HAK cv_HAK 0001 1365 74 19 71 0002 206 35 31 48 biomass_by_species_trip Optional table presented at the end of the output from a multi trip analysis containing all biomass estimates including those for sub populations but not stratum biomasses This table is s
13. projected_cvs strata 0002 0003 0004 extra_stations 1 2 3 and run SurvCalc b gt myfile As well as calculating all the phase 1 biomass estimates and c v s this run will produce a table like species population projected_cv HOK all 1 83 HAK all 25 7 giving estimates of the likely c v s at the end of the survey if we were to use the hoki allocation We can repeat this run changing the allocation in the projected_cvs command block until by trial and error we find the most suitable phase 2 allocation Another reason for considering alternative phase 2 allocations is that some strata may be more expensive in terms of steaming time to return to than others Suppose for example that in the above example stratum 0002 was so far away that including it in phase 2 would mean that there was time for only five phase 2 stations Then we could use projected_cvs to compare the projected c v s from the following two alternative allocations phase 2 allocation stratum with str 0002 without str 0002 0001 0 0 0002 1 0 0003 2 2 0004 2 3 0005 0 ai Total 5 6 The first of these allocations is what we get from the above table of gains if we allocate only five phase 2 stations and the second is the result of allocating six phase 2 stations but ignoring stratum 0002 Some practical considerations In theory the rule is that our phase 2 calculations using either task phase2_calc or task calc_biomass with command projected_
14. variance mean_squared Defines which algorithm will be used in calculating gain see Section 5 5 This sub command is included only for completeness Users should omit it unless they have a very good reason to use the variance algorithm Maximum number of additional stations per stratum phase_2 Positive integer None Defines the maximum number of additional stations in each stratum for which gains will be calculated i e the number of columns in the output table of gains see Section 5 5 23 projected_cvs Conditions Effects Example strata Command Conditions Type Effects extra_stations Command Type Effects Calculate projected biomass c v s during a survey Ignored unless run_time task is calc_biomass or calc_biomass_and_LFs Can only be used in a single trip analysis Calculates the projected biomass c v s 1 e the c v s expected at the end of the survey given the data to date see Section 5 6 Defines any following subcommands as being projected_cvs subcommands The following command block could be used during a 2 phase survey at the time when all stations had been completed except for 2 in stratum 002A and 3 in stratum 002B projected_cvs strata 002A 002B extra_stations 2 3 Strata that are not yet completed projected_cvs There must already be at least two stations in all strata specified here String vector of stratum names Specifies which strata are not yet completed i e those in
15. vector trips Defines any following subcommands as being new_st rata subcommands for the specified trip Omit the label if only one trip is being analysed i e if trips is of length 1 Areal availabilities for the new strata will be assumed to be 1 Different values which are trip and species specific may be set using command areal_availability The following command creates new strata 003A and 003B for trip tan0601 with areas 2153 and 397 respectively and areal availabilities 1 and 0 8 respectively new_strata tan0601 strata 003A 003B areas 2153 397 Names of new strata new_strata Must be different from the names of existing strata String vector Defines the names of new strata A warning should be output if any string in names does not occur in either change_strata trip toorreassign_strata trip strata Area km of each new stratum new_strata A numeric vector of the same length as names Defines the areas of new strata Change the areas of existing strata A trip code must be in the vector trips Defines any following subcommands as being change_stratum_area subcommands for the specified trip Changes the area of the strata with names in names to the values in new_areas Omit the label if only one trip is being analysed i e if trips is of length 1 The following command block changes the area of strata 004A and 004B to 3152 and 793 respectively change_stratum_area tan0601 strata 004A 004B new_areas 31
16. which there are still some more stations to be done Number of additional stations per stratum projected_cvs Constant vector of same length as strata Specifies how many more stations are to be done in each stratum named in subcommand strata 24 3 1 5 Commands defining output The main output file produced by SurvCalc contains obligatory and optional elements The commands in this section together with command station_data_columns in Section 3 1 1 allow the user to specify which optional elements should be included They also specify to what precision certain numbers in both the main and other output files should be given output_tables Label Conditions Effects Notes Examples Tables to include in the main output file None Ignored unless biomass or LFs are being calculated i e the task is calc_biomass calc_biomass_and_LFs or calc_LFs Defines any following subcommands as being output_tables subcommands See Section 4 1 1 for descriptions of all obligatory and optional tables of biomass and LFs The following example requests tables of sub population biomass and LFs by stratum but not LFs by station output_tables sub_biomass_by_stratum T LFs_by_stratum T LFs_by_station F sub_biomass_by_stratum Output table s of sub population biomass estimates and c v s by stratum Command output_tables Conditions Ignored unless biomass is being calculated i e the task is calc_biomass or calc_biomas
17. 0 3 1 4 Commands defining the calculations The commands in this section describe the user s preference for various options in the calculations There are options associated with fish density see Section 5 1 4 sub populations defined by sex and or length ranges see Section 5 3 1 and the scaling of LFs see Section 5 4 preferences Label Conditions Effects Notes distance_towed Command Conditions Type Default Effects Notes width_swept Command Conditions Type Default Effects Notes catch_weight Command Type Default Effects Notes User preferences for fish density calculations A trip code must be in the vector trips Ignored if the task is output_lw_coeff or if it is calc_LFs and LF_scaling is not numbers_per_km2 or if the user has provided station catch data W Defines any following subcommands as being preferences subcommands for the specified trip Omit the label if only one trip is being analysed i e if trips is of length 1 User preference for calculation of distance towed at each station preferences Must not be used when survey_type pot String vector containing one or more of the following options in order of preference recorded_distance recorded_speed time constant_speed time from_lat_lon None Defines the user preference for the method or methods in priority order of defining the distance towed at each station Any station for which distance towed
18. 16 14 In this example the six biggest numbers in the table have been underlined to help illustrate how the table of gains should be interpreted This shows us that the optimum allocation for 6 phase 2 stations is one station in stratum 0002 two stations in stratum 0003 and 3 stations in stratum 0004 If it turned out that there was time only for five phase 2 stations then we should ignore the smallest underlined number and thus do only two extra stations in stratum 0003 The third way that SurvCalc can be useful in 2 phase surveys is in evaluating alternative phase 2 allocations using the command projected_cvs If for example hake was also a target species in the above survey we would need to add the command block phase_2 HAK n_stations 6 to our main input file before running SurvCalc 2 gt myfile Then myfile would contain two tables of gains one for hoki and one for hake from which we could construct a table like the following showing two alternative phase 2 allocations phase 2 allocation stratum for HOK for HAK 0001 0 2 0002 UB 2 0003 2 1 0004 3 1 0005 0 0 So which phase 2 allocation should we use the one for hoki the one for hake or some compromise between the two To answer this question the first thing to do is to find out 49 what c v s we might expect to get for the hoki and hake biomass estimates using each of the above allocations To evaluate the hoki allocation we would add the command block
19. 52 793 Names of strata whose areas are to be changed change_stratum_area Must be names of existing strata String vector New areas for strata whose areas are to be changed change_stratum_area Numerical vector 15 3 1 3 Commands extending the data The commands in this section allow the user to extend the data to be analysed by providing length weight coefficients or setting the multiplicative factors which affect either the calculation of fish density vulnerability vertical availability and area fished see Section 5 2 or the calculation of biomass from fish density areal availability or population area see Section 5 3 Length weight coefficients need be provided only if they are needed see Section 5 1 6 and if the default values in the database are not present or correct Most of the multiplicative factors have default values 1 for vulnerability and both areal and vertical availability the stratum area for population area Other values must be provided separately for each combination of trip and species Note that vulnerability vertical availability and area fished are associated with stations areal availability and population area are associated with strata vulnerability Vulnerability of a species to capture at each station in a trip Label trip_code _ species_code Conditions Ignored if the user has provided station catch data W Must not be used when survey_type pot Effects Defines any following subco
20. Cie is taken directly from column weight in table t_subcatch so SurvCalc must read columns trip_code station_no species subcatch_no and weight from this table With option calculated Es 0 0010_ N W P Where ae means the sum over all sexes k and all length ijk ijk records j in station i for which subcatch_no c 48 6 SurvCalc AND 2 PHASE SURVEYS SurvCalc and the associated R library can be used in three ways to help in 2 phase surveys First the R function allocate can be used in designing the survey see Appendix 2 One aspect of the design is the percentage of stations to be done in phase 1 of the survey Note that although Francis 1984 recommended this be 75 the current recommendation based on much more extensive simulations is 90 Francis 2006 Second task phase2_calc of SurvCalc can be used at the end of phase 1 to help in determining the allocation of phase 2 stations to strata For example suppose our target species is hoki and at the end of phase 1 we decide there is sufficient remaining survey time to complete six phase 2 stations Then we could insert the command block phase_2 HOK n_stations 6 into our main input file and run SurvCalc 2 gt myfile Within the output file my file there will be a table of gains for hoki which might look like stratum pi 2 3 4 5 6 0001 19 15 12 10 8 7 0002 62 44 33 26 21 17 0003 100 60 40 29 21 17 0004 96 64 46 34 21 2l 0005 43 32 25 20
21. E iba celeb enediedene 37 4 4 Precision of numbers in output files ee eee esse ceseceeeeeeeeeseeeeeeeeaeeeseeeaeeeaaeenaes 38 CALCULATIONS IN SurvCalunaitonisicas iii a lente brad 39 5 1 Data and Nottm augusta 39 5 1 1 Note Omsubcatches iii ideada ideada 41 5 1 2 Note on stations and strata without LF data ooooocnncnnccinccoccnnononancnancnnnncnnncnnnos 41 5 1 3 Excluding stations and strata 0 see eesecssecesecsseceeceseeeeeeeseeeseeeeaeeeseeeaaeeaaeenaes 41 5 1 4 User preferences for fish density variables 2 0 0 0 cesceeceeeeeeeeeereeeeeeneeeneeenaes 42 5 1 5 Calculation ENS carita 42 5 1 6 Use of length weight coefficientS oooonnncnnnnnnnncnoccnoncconcconnnonnconncnnna nac crac nrna conan 42 5 2 Calculating fish Gensittes viiuriooiinncodorica corria nena sinesine ini as 43 5 3 Calculating DIOMASSES sceoccaionrionenionrnacinnrio nece nosuconddenadoo be sant sa na E Eaa ei aans 43 5 3 1 Calculating sub population biomasses oooonnnccnononocanoncnonanonnnonn conc conc cono nrna cra 43 A ON 5 4 Calcolatine LE ds 44 5 4 1 Calculatinsicvis for BES unos tects eaae EE E EE EERE E E ie 45 5 5 Calculating phase 2 Gains trio ne a E E EE A E a a a as 46 5 6 Calculating projected Vii a eens aa 47 5 7 Output forcatch at a ges ii AAA A A E EEE a 48 SurvCalc and 2 PHASE SURVEYS croio E ea i E E nono E eta aeaa Nee 49 REFERENCES curar ta 51 Appendix 1 Command block forMat oonccnoncnoncconcconnconnnnnnnnnnconnnnnn nono nono nono
22. Effects Notes Examples quantity Command Type type Command Type precision Command Type Precision of quantities in the output files None Ignored when the task is out put_LW_coeffs Defines any following subcommands as being output_precision subcommands Use this command only if the default level of precision as given in the following example is unsatisfactory This command affects both the main output file and files station out station_catch out and stratum_catch out see Section 4 4 for more details The following example shows the default level and type of precision for all output quantities output_precision quantity density biomass LF_number cv gain type sig_fig sig_fig sig_fig dec_place dec_place precision 4 4 4 1 0 Type of quantity for which precision is being specified output_precision String vector containing any or all of density biomass LF_number cv gain Type of precision being specified significant figures or decimal places output_precision String vector of same length as quantity containing either sig_fig or dec_place at each position Number of significant figures or decimal places for each quantity output_precision Integer vector of same length as quantity 3 1 6 Examples of main input files One simple way of creating a valid main input file is to edit an existing file With that in mind here are three example files of increasing complexity All could be used with tasks
23. a The remainder of this section compares SurvCalc with its predecessor trawlsurvey and discusses some possible future extensions of SurvCalc Sections 2 describes how to run SurvCalc Sections 3 and 4 describe the various input and output files respectively Section 5 documents the calculations in SurvCalc and Section 6 discusses how SurvCalc output should be used in 2 phase surveys 1 2 Relationship of SurvCalc to trawlsurvey This section is aimed at past users of the program trawlsurvey Vignaux 1994 and may safely be ignored by others It is intended to help introduce these past users to the main features of SurvCalc by comparing it with the earlier program From the user s point of view a major difference between trawlsurvey and SurvCalc is the way in which they define the analyses they want For trawlsurvey this was done by entering information via a series of blue screens whereas for SurvCalc this information is written into the main input file in a command block format similar to that for CASAL This file together with the SurvCalc manual will serve as a complete documentation of the analysis A second important difference concerns the computers on which each program will run and their data requirements trawlsurvey runs only on Unix machines and requires the survey data to be in an Empress database structured like trawl Mackay 2000 In contrast SurvCalc runs on both Unix and Windows machines and can access survey data either
24. a full point There is no need to mark the end of a command block This is automatically recognised by either the end of the file or the start of the next command block which is marked by the on the first character of a line Also note that the commands sub commands and arguments in the parameter files are case sensitive Some commands can never have subcommands such as survey_type If a command has no subcommands then it has to have arguments which are placed on the same line as the command All other commands have no arguments but have subcommands instead Some commands can be used multiple times and if they are they must have a different label each time see Table 2 52 9 APPENDIX 2 THE SURVCALC R LIBRARY This library can be downloaded from https one niwa co nz display FISHMOD SurvCalc It contains two main functions input SurvCalc and allocate input SurvCalc Function input SurvCalc simply reads the main output file of SurvCalc into R Thus if you have a SurvCalc output file myout in directory H proj hok you can read this into the R object hokout using the R command hokout lt input SurvCalc H proj hok myout or alternatively hokout lt input SurvCalc myout H proj hok The object hokout is a list of lists that is structured hierarchically by trips and then species For example all the output concerning trip buc8501 in myout is contained in the list hokout buc8501 To
25. a survey or afterwards Each can be applied to analyse multiple species in multiple surveys or trips in a single run of SurvCalc and the species analysed may be different in different trips 1 Task calc_biomass Calculates biomasses by stratum and overall Can also calculate biomasses for sub populations defined by sex and or length range e g for males of length between 20 cm and 80 cm C v s are calculated for all biomasses Optionally calculates during a survey projected biomass c v s 1 e the c v s expected at the end of the survey given the data to date this can be useful during a 2 phase survey 2 Task calc_LFs Calculates LFs by station and or stratum and or overall All LFs are presented by sex including a category for unsexed and overall The user can choose between five alternative methods of scaling the LFs C v s are not calculated for LFs 3 Task calc_biomass_and LFs Combination of tasks calc_biomass and calc_LFs but only one method of LF scaling is allowed scaling to represent estimated numbers in the population and c v s are optionally calculated for LFs by stratum and sex The next task will usually be used at sea at or near the end of phase 1 of a 2 phase survey Francis 1984 It can be applied only to a single trip but can involve multiple species and is intended to provide information useful in deciding on the phase 2 allocation 1 e how a specified number of phase 2 stations should be allocated among
26. able SurvCalc automatically constructs an SQL to extract data from table t_stratum and this ends with a where clause restricting this extraction to the specified trip s SurvCalc will append t_stratum_where to this where clause using and Restrictions for extracting catch data where String must be a valid SQL Boolean expression Specifies criteria to restrict the selection catch data from t_catch table SurvCalc automatically constructs an SQL to extract data from table t_catch and this ends with a where clause restricting this extraction to the specified trip s and species SurvCalc will append t_catch_where to this where clause using and Restrictions for extracting length data where String must be a valid SQL Boolean expression Specifies criteria to restrict the selection catch data from t_1gth table SurvCalc automatically constructs an SQL to extract data from table t_lgth and this ends with a where clause restricting this extraction to the specified trip s and species SurvCalc will append t_lgth_where to this where clause using and 13 3 1 2 Commands modifying the data The commands in this section allow the user to modify the stratification in the data that have been read in either from a database or flat files The following changes can be made stations can be reassigned to different strata either existing ones or new user defined strata and areas of exis
27. ach sub population Each table of gains will have n rows i e one per stratum and phase_2 n_station columns Let G be the relative gain i e the estimated reduction in the biomass variance associated with adding the eth extra station to the sth stratum To calculate G we need first to calculate AD and ad if sub populations are involved and its s e using equations 1 3 and 5 The formula for G depends on whether phase_2 algorithm is mean_squared or variance 46 a AD Y mean_squared n e 1 n e 30 a s e AD n A variance n e 1 n e for sub populations replace AD by ad in this equation Before being output each table of gains is to have maximum value 100 using the equation G 100G max s G 31 5 6 Calculating projected c v s Projected c v s are calculated for all biomasses including those for sub populations when the task is calc_biomass or calc_biomass_and_LFs and there is a projected_cvs command block see Section 3 1 4 in the main input file In these calculations n is the number of stations already completed in stratum s and e is the number of stations yet to be completed the latter being specified in the projected_cvs command block For the projections it is assumed that in each stratum the mean and variance of the catch rates for the remaining stations will be exactly the same as for the existing stations The calculations use the same equations as are
28. analysed which stations to use from that trip which database if any should be accessed and what additional data should be extracted Commands modifying the data Section 3 1 2 What changes should be made to the data extracted from the database stations can be reassigned to different strata either existing ones or new user defined strata areas of existing strata can be changed Commands extending the data Section 3 1 3 Non database information needed for the analyses vulnerability and vertical availability for each station areal availability for each stratum area fished for pot surveys length weight coefficients 10 Commands defining the calculation Section 3 1 4 User s preference for various options in the calculations how should the distance towed at each station be calculated from start and finish positions or from recorded speed and time etc what should be the width swept at each station the recorded doorspread or a specified constant what sub populations if any should biomass be calculated for how should LFs be scaled information needed for phase 2 calculations and projected c v s Commands defining output Section 3 1 5 What tables should be included in the output file and what degree of precision should be used for different categories of output fish density biomass LF numbers c v gain Some examples of main input files are given in Section 3 1 6 Commands that may be and sometimes must
29. ations command for HOK will apply to all trips for which this species 1s to be analysed trips tan0617 tan0714 species tan0617 codes HOK HAK LIN species tan0714 codes HOK HAK A input_from_database database Empress where tan0617 t_station gear_perf lt 3 and station_no 60 where tan0714 t_station gear_perf lt 3 preferences tan0617 distance_towed recorded_distance recorded_speed tim width_swept recorded_doorspread catch_weight recorded calculated preferences tan0714 from_lat_long distance_towed recorded_distance recorded_speed tim width_swept recorded_doorspread catch_weight recorded vulnerability tan0617_HOK default_value 1 other_stations 71 72 other_values 0 8 0 8 29 from_lat_long sub_populations HOK sexes all all Lmin 0 60 Lmax 59 200 labels to60 60 output_tables sub_biomass_by_stratum T biomass_by_species F biomass_by_species_stratum F LFs_by_stratum T LFs_by_station F Number_measured T LF_totals T output_precision quantity density biomass LF_number type dec_place dec_place sig_fig precision 0 0 8 30 cv dec_place 0 gain dec_place 1 3 1 7 Repeated commands in the main input file The examples in Section 3 1 6 have shown that some commands can be repeated within the main input file When they are repeated their labels show exactly what each command applies to either a trip a species or a combination of a trip and spe
30. calculated A species code must be in species trip code for some trip in trips Ignored unless the task is one of calc_biomass_and_LFs orphase_2_calc Defines any following subcommands as being sub_populations subcommands for the specified species For analyses covering more than one trip this command will affect the output for the given species in all trips where that species is analysed The following command requests ORH biomass estimates for six sub populations and supplies labels to be used for these in the output file sub_populations ORH calc_biomass sexes all all male mal female femal Lmin 0 20 0 20 0 20 Lmax 20 80 20 80 20 80 labels A_sm A_lg M_sm M_lg F_sm F_lg Sexes of sex based sub populations sub_populations String vector containing one or more of all male female and unsexed Specifies the sex of each sub population for the given species Length bounds for each length based sub population sub_populations Numeric vector of same length as sexes Specifies length bounds for each sub population for the given species Sub population includes all fish with Lmin lt length lt Lmax User supplied labels for sex based sub populations sub_populations String vector of same length as sexes Supplies labels for sub population biomass estimates in the output Vessel speed to be used in calculating distance towed A trip code must be in the vector trips Ignored if the user has provided sta
31. can not be defined will be dropped from the analysis See Section 5 1 4 for more details about the options for distance_towed User preference for defining width swept at each station preferences Must not be used when survey_type pot String vector containing one or more of the following options in order of preference recorded_doorspread constant_doorspread recorded_wingspread None Defines the user preference for the method or methods in priority order of defining the width swept at each station Any station for which width swept can not be defined will be dropped from the analysis See Section 5 1 4 for more details about the options for width_swept User preference for calculation of catch weight at each station preferences String vector containing one or more of the following options in order of preference recorded calculated None Defines the user preference for the method or methods in priority order of defining the catch weight at each station Any station for which catch weight can not be defined will be dropped from the analysis See Section 5 1 4 for more details about the options for cat ch_weight 21 sub_populations Label Conditions Effects Notes Examples sexes Command Type Effects Lmin Lmax Command Type Effects Notes labels Command Type Effects constant_speed Label Conditions Effects Notes value Command Type Notes sub populations for which biomasses should be
32. catch file A file of station catch data is output with x switch in the command line if run time task is one of output_flat_files calc_biomass calc_LFs or calc_biomass_and_LFs The file contains all the columns names in the station table plus three columns for each of the species analysed as well as two columns for each of the sub_populations defined by the user For species XYZ these columns are XYZ_kg the catch weight C XYZ_kg_km the linear density LD and XYZ_kg km2 the areal density AD For sub_population ABC of species XYZ the two columns are XYZ_ABC_kg the catch weight c and XYZ_ABC_kg_km2 the areal density ad For stations from trips in which species XYZ is not analyzed NULL is entered for all columns associated with this species 36 4 2 2 Output to stratum catch file A file of stratum catch data is output with y switch in the command line if run time task is one of output_flat_files calc_biomass calc_LFs or calc_biomass_and_LFs The file contains columns trip_code stratum and area_km2 plus three columns for each of the species analysed For species XYZ these columns are XYZ_kg stratum mean catch weight 2 C n XYZ_kg_km the linear density LD and XYZ_kg km2 the areal density AD For strata from trips in which species XYZ is not analyzed NULL is entered for all columns associated with this species 4 3 Catch at age data output The command SurvCalc lt gt myout run produces
33. cies Some of these commands must be repeated e g preferences must be repeated for each trip and others need not be because there is a default action when they are not repeated e g when vulnerability is not repeated for a trip species it is assumed to be 1 for all stations for that trip species The requirements for all commands that may be repeated are summarised in Table 2 Labels may be omitted from these commands when this causes no ambiguity For example those commands that are usually labelled by trip need no label a when they are intended to apply to all trips or b in a main input file that concerns only one trip Similarly if the file concerns only one trip and one species then the commands normally labelled by trip and species do not need a label Table 2 Requirements for all commands that can be repeated in a main input file Must be Default action if Command Label repeated command not repeated species trip Yes preferences trip Yes where trip No Use default selects change_strata trip No No changes reassign_strata trip No No changes new_strata trip No No new strata change_stratum_area trip No No changes constant_speed trip Maybe constant_doorspread trip Maybe sub_populations species No No sub populations phase_2 species Yes vulnerability trip amp species No All 1 vertical_availability trip amp species No All 1 areal_availability trip amp species No All 1 population_area trip amp s
34. cvs should be done only after we have finished all phase 1 stations and before we have started phase 2 However it is perfectly acceptable to break this rule in some situations Remote strata provide a good reason to do phase 2 calculations before the end of phase 1 After completing the phase 1 stations in a remote stratum we might want to use task phase2_calc to work out on the basis of the phase 1 stations completed to date whether we are likely to need any phase 2 stations in that stratum and if so approximately how many If it looks like some phase 2 stations will be needed in this stratum it is quite permissible to do them immediately before leaving the area 50 Another reason to break the rule is if part way through phase 2 it becomes clear that the actual number of phase 2 stations will differ from what was assumed so that there is a need to modify the original phase 2 allocation There is one rule about phase 2 calculations that should never be broken That is you must never include catches from phase 2 stations when doing phase 2 calculations Suppose for example that during phase 1 you added two phase 2 stations numbers 32 and 33 say in a remote stratum Then these stations must be excluded from any subsequent phase 2 calculations This could be done using subcommand t_station station_no not in 32 33 of command where If you were already using this subcommand say to exclude stations with poor gear performance then th
35. d strata including their effective areas and one of the selected stations including the stratum associated with each The PREFERENCES is simply a repeat of the information in command preferences The results for a particular species and trip always starts with a list of exclusions which stations and strata are excluded and why and summaries of the numbers of stations and strata that were selected and used After this the remaining output is in a series of tables some obligatory and others optional the latter being requested with command output_tables 33 4 1 1 Tables in the main output file area_swept_factors Obligatory table when area swept is calculated produced for each species trip combination Summary statistics for the two factors that define area swept mean min max width_swept_m 120 56 104 FA Os caf distance_towed_km 5 47 3 82 5 76 stratum_summary Obligatory table when biomasses are being calculated produced for each species trip combination One line per stratum and including mean fish densities biomass and c v NtowsHOK is the number of tows with non zero catch for HOK or whatever the specified species is The area is the population area a which may differ from the stratum area if commands areal_availability or population_area are used stratum area Ntows NtowsHOK kg km2 biomass cv biomass 0001 2150 4 3 9 19 71 0002 1318 4 4 24 31 48 biomass Obligatory table when biomasses are being calculated
36. e AD from columns with names like XYZ_kg_km2 LD from columns with names like XYZ_kg km Otherwise they are calculated as AD 10 C f 1 LD 1000C 1 852d u v 2 EW d Stratum mean densities are calculated as AD 2 AD n 3 LD Y LD n 4 The s e s of areal densities are given by E ap aD y Y ln 1 017 5 s e AD ll 5 3 Calculating biomasses Stratum and overall biomasses are calculated as B AD a 1000 6 B B 7 and their standard errors are calculated as s e B s e AD a 1000 8 Y se 8 9 n O ws II 5 3 1 Calculating sub population biomasses The first step in calculating sub population biomasses for each stratum is the calculation of sub population catches c for each station In the special case of a stratum with zero catch 1 e 2 C 0 the sub population catch is set to 0 for all stations in that stratum For all other strata a check is done to see whether there are enough LF data to warrant the calculation of sub population catches This involves the calculation of the quantity 43 eS 10 which is the proportion of the catch in stratum s that is from stations with LF data Sub population biomass estimates will be unreliable in strata where p is low Any strata with Ps lt 0 3 are dropped from the calculation of sub population biomasses and a warning message is included in the output file Strata with 0 3 lt p
37. eed constant_speed x t Wi width_swept recorded_doorspread dist_doors in t_station constant_doorspread constant_doorspread recorded_wingspread dist_wings in t_station C catch_weight recorded weight in t_catch calculated C 0 001 N Wy J P Note that if there is no record in t_catch for the given species station and trip this means that C 0 5 1 5 Calculation of c v s All c v s are calculated and presented as percentages Unless otherwise stated the c v of any variable X is calculated as c v X 100s e X X 5 1 6 Use of length weight coefficients Length weight coefficients aj by Cx are not always required They are used only in three situations when C is calculated rather than being a recorded value Table 6 when sub population biomasses are calculated see equation 11 Section 5 3 1 and when LFs are calculated and scaled to numbers in population see calculation of CF and CF in Section 5 4 42 5 2 Calculating fish densities One or both of two types of densities may be calculated areal kg km2 or linear kg km The former are always used for calculating either biomass or gains The latter not defined if survey_type is pot are calculated only if requested with commands station_data_columns and or stratum_data_columns as output columns in station or stratum tables When a station catch input file is used see Section 3 2 the station densities are read directly from this fil
38. el Effects Notes Example t_station Command Type Effects Notes t_stratum Command Type Effects Notes t_catch Command Type Effects Notes t_lgth Command Type Effects Notes Restrict the selects from the database tables A trip code must be in the vector trips Restricts the records selected from one or more of database tables t_stratum t_station t_catch and t_lgth Defines any following subcommands as being where subcommands for the specified trip Most users will want to use only subcommand t_station to define the station select The label may be omitted if only one trip is being analysed 1 e if trips is of length 1 The following command block restricts the stations selected to those for which gear_perf is less than 3 and station_no is less than 100 where t_station gear_perf lt 3 and station_no lt 100 Restrictions for extracting station data where String must be a valid SQL Boolean expression Specifies criteria to restrict the selection of station data from t_station table SurvCalc automatically constructs an SQL to extract data from table t_station and this ends with a where clause restricting this extraction to the specified trip s SurvCalc will append t__st ation to this where clause using and Restrictions for extracting stratum data where String must be a valid SQL Boolean expression Specifies criteria to restrict the selection stratum data from t_stratum t
39. equired because the command sub_populations requires a label a species code when multiple species are analysed No sub population biomasses will be calculated for the other species HAK and LIN unless an additional sub_populations command block or blocks is included with the appropriate label Note also that because of its label the vulnerability gain dec_place 1 command block has no effect on the biomass estimates for HAK and LIN trips tan0617 species tan0617 codes HOK HAK LIN input_from_database database Empress where t_station gear_perf lt 3 and station_no 60 preferences distance_towed recorded_distance recorded_speed tim width_swept recorded_doorspread catch_weight recorded calculated vulnerability tan0617_HOK default_value 1 other_stations 71 72 other_values 0 8 0 8 28 from_lat_long sub_populations HOK sexes all all Lmin 0 60 Lmax 59 200 labels to60 60 output_tables sub_biomass_by_stratum T biomass_by_species F biomass_by_species_stratum F LFs_by_stratum T LFs_by_station F Number_measured T LF_totals T output_precision quantity density biomass LF_number cv gain type dec_place dec_place sig_fig dec_place dec_place precision 0 0 8 0 1 Example 3 The final example extends the second example to analyse multiple trips Note that for each trip a list of species to be analysed a where command and a set of preferences is provided The sub_popul
40. es in all strata in the given trip except for those in other_strata Strata in which the areal availability differs from the default value areal_availability Must be existing strata for the given trip String Specifies strata in which the areal availability differs from the default value Areal availability values that differ from the default value areal_availability Positive numeric vector of same length as other_strata Specifies the areal availabilities for those stratain other_strata 18 population_area Label Conditions Effects Notes Examples other_strata Command Conditions Type Effects other_values Command Type Effects lw_coeff Label Conditions Effects Notes Examples a b c Command Conditions Type Default Population area for species in each stratum in a trip trip_code _ species_code For each combination of trip and species to be analysed there must not be both an areal_availability command block anda population_area command block use one or the other or neither Defines any following subcommands as being population_area subcommands Specifies a population area for the given species at all existing strata in the given trip This is used to calculate areal availability which is population area divided by stratum area Each population_area command block applies to one species in one trip The command may be omitted when the population area of the species is the same as the s
41. f strata with LF samples or zero catch LF correction factor i indexes the selected stations for the specified trip ti di C ci Wi Vi Uyi fi AD adi LD time towed h distance towed n mile catch weight kg width of tow m vulnerability vertical availability effective area fished m3 areal density kg km linear density kg km s indexes the selected strata for the specified trip as As s AD ad LD AD s LF Ps B bs CF K stratum area km number of selected stations in stratum s number of selected stations in stratum s where there is LF data or C 0 areal availability population area km mean areal density kg km mean linear density kg km mean areal density kg km over stations with LF data or C 0 proportion of catch taken in stations with LF data biomass t LF correction factor LF scaling factor k indexes sex 1 male 2 female 3 unsexed 4 all ak Diy Ck length weight coefficients calculates weight in g from length in cm 39 equation 24 calculated from time_s time_f date_s date_f Table 6 Table 6 equations 10 11 Table 6 from vulnerability from vertical_availability f 1852dwyyu or Oarea_fished equation 1 or station data file equation 2 area_km2 from areal_availability or stratum data file a a fu equation 3 equation 4 equation 19 equation 10 equation 6
42. fault_value Command Type Default Effects other_stations Command Conditions Type Effects other_values Command Type Effects area_fished Label Conditions Effects Notes Examples default_value Command Type Default Effects trip_code _ species_code Ignored if the user has provided station catch data W Must not be used when survey_type pot Defines any following subcommands as being vertical_availability subcommands Specifies a vertical availability for the given species at all existing stations in the given trip Each vertical_availability command block applies to one species in one trip The command may be omitted when the vertical availability of the species is at all stations in the trip The following command block specifies that the vertical availability of HOK in trip tan0601 is 0 8 and 1 2 for stations 33 and 35 respectively and 1 for all other stations vertical_availability tan0601_HOK default_value 1 other_stations 33 35 other_values 0 8 1 2 Default value for vertical availability of the given species in the given trip vertical_availability Positive number 1 Defines the vertical availability of the given species in all stations in the given trip except for those in other_stations Stations at which the vertical availability differs from the default value vertical_availability Must be existing stations for the given trip Numeric vector Specifies stations at w
43. g km are chosen by the user although the biomass estimates are always based on densities in kg km This is potentially misleading because the obvious inference from this table is that the biomass estimates derive from the presented densities In SurvCalc the densities in this table are always in kg km but kg km can be calculated if requested and output to separate station catch and or stratum catch files 1 2 3 Corrections to trawlsurvey 1 LFs for stations and species with more than one subcatch are not well handled in trawlsurvey For such stations there can be more than one record in the trawl table t_lgth with the same station_no species and lgth as in the following extract trip_code station_no species subcatch_no Igth percent_samp no_a tan0601 79 HOK 1 56 5 53 2 tan0601 79 HOK 2 56 10 55 8 In trawlsurvey this will misleadingly produce two lines for length 56 cm in the LF for station 79 but there s no problem with stratum and overall LFs This does not occur with SurvCalc 2 Although the trawlsurvey blue screen says that there should be no overlap in the length ranges of sub populations the program actually does allows overlap and sometimes this causes errors e g when accidentally two identical length ranges were specified trawlsurvey produced just one sub population biomass for this length range but this biomass was too high by a factor of 2 SurvCalc allows overlap in length ranges and treats these co
44. hich the vertical availability differs from the default value Vertical availability values that differ from the default value vertical_availability Positive numeric vector of same length as other_stations Specifies the vertical availabilities for those stations in other_stations Area fished m for a species at each station in a potting survey trip_code _ species_code Must not be used except when survey_type pot Defines any following subcommands as being area_fished subcommands Specifies an area fished for the given species at all existing stations in the given trip Each area_fished command block applies to one species in one trip The following command block specifies that the area fished for BCO in trip abc0601 is 27 m and 25 m for stations 33 and 35 respectively and 30 m for all other stations area_fished abc0601_BCO default_value 30 other_stations 33 35 other_values 27 25 Default value for area fished for the given species in the given trip area_fished Positive number None Defines the area fished for the given species at all stations in the given trip except for those in other_stations 17 other_stations Command Conditions Type Effects other_values Command Type Effects areal_availability Label Conditions Effects Notes Examples default_value Command Type Default Effects other_strata Command Conditions Type Effects other_values Command Type Effects Stati
45. iation in numbers between stations in the same stratum All other sources of uncertainty are ignored Thus because we can write and we ignore uncertainty in K and CF c v TN c v MTN which leads to equation 25 Similarly we can write TN MTN K a cr 29 s LF but we ignore uncertainty in both the K and the term in square brackets Thus c v 7N c v MIN K which leads to equation 26 5 5 Calculating phase 2 gains Gains are calculated only when the task is phase_2_calc and the calculations are based on the equations in Francis 1984 When these calculations are done it is assumed that the selected stations i e those defined by commands in Section 3 1 1 or included in the station flat file if there is one are all the phase 1 stations in a 2 phase survey Sometimes for logistical reasons the calculations will be done before all phase 1 stations are completed and or after some phase 2 stations have been done In this situation the calculations should use all completed phase 1 stations but not any of the phase 2 stations to use phase 2 stations in the calculations would effectively change the survey from 2 phase to multi phase and the statistical behaviour of multi phase surveys of this type are not known A table of gains will be calculated for each species in phase_2 species If sub populations are defined for any of these species then an additional table of gains will be calculated for e
46. iles that need to be provided depend on the run time task and some of the preferences specified in the input file e g biomass calculation involving no sub populations and using the recorded catch preference requires stratum S station T and catch data files U or alternatively stratum S and station catch X data files SurvCalc will give error messages if data provided are inconsistent with the run time task and preferences 3 INPUT FILE SPECIFICATIONS SurvCalc requires a main input file described in Section 3 1 which describes the data that are to be analysed and specifies some details of the analyses and the desired output The actual data to be analysed are read either from an Empress or PostgreSQL database or from user provided flat files described in Section 3 2 3 1 The main input file The main input file for SurvCalc has default name input sic but this name can be changed if run time argument f is used see Section 2 It uses a command block format similar to that used in CASAL The order of command blocks within the main input file and of subcommands within a command block is arbitrary other conventions of this format are described in Appendix 1 The commands used in the main input file fall into five categories depending on their function as follows Commands defining the data Section 3 1 1 What data are to be analysed what type of survey trawl or pot which trip and species are to be
47. imply the union of all the biomass tables ordered by species and then by trip within species species trip population lbound biomass ubound cv Nstations HOK tan0601 all 31601 46081 60561 16 97 HOK tan0601 to60 114 846 1807 57 90 HOK tan0601 60 11830 20001 28172 20 90 HOK tan0701 all 29701 51315 63517 18 95 projected_cvs Obligatory table when biomasses are calculated and command projected_cvs is used Includes projected c v s for all total biomasses including sub populations if any species population projected_cv HOK all 10 HOK to60 23 HOK 60 11 HAK all 16 LF _correction_factors Obligatory table for task calc_biomass_and_LFs produced for each species trip combination see CF and CF in Section 5 4 stratum correction_factor 0001 1 08 0002 1 08 all 1 04 LF_overall Obligatory table when LFs are calculated produced for each species trip combination A single table for the whole population as follows N means number measured SN means scaled number L Nfemale Nmale Nunsexed Nall SNfemale SNmale SNunsexed SNall 32 1 3 0 4 2199 1 1783 3 0 3983 2 33 8 5 0 13 9259 1 16399 2 O 25658 3 LFs_by_stratum Optional table produced for each species trip combination Same format as LF_overall but with an additional initial column stratum 35 LFs_by_station Optional table produced for each species trip combination Same format as LF_overall but with an additional initial columns stratum and stati
48. ion 5 7 This attribute extracted only for task output_for_catch_at_age see Section 5 7 survey_type Type of survey Type String must be either trawl or pot Default trawl Effects Determines which alternative variables and equations will be used in calculations and what other input file commands are valid Other types of survey that may be allowed in future versions of SurvCalc include scallop and oyster Notes This command is not needed for trawl surveys trips The trip or trips that should be analysed Type String vector each member of which must be a valid trip code Effects Limits the data extracted from the database to that relating to the specified trip or trips 11 species Label Effects Notes codes Command Conditions Type Effects The species to be analysed for a specified trip A trip code must be in the vector trips Defines any following subcommands as being species subcommands for the specified trip Omit the label if only one trip is being analysed i e if trips is of length 1 Species codes species Only one species can be analysed i e codes must have length 1 if the task is output_for_catch_at_age String vector each member of which must be a valid species code Limits the catch and or length data extracted from the database for the specified trip to that relating to the specified species input_from_database The interface to the database to extract the data Effects Notes da
49. ion for excluding a stratum depends on the task Stratum s is excluded if n lt 2 for tasks calc_biomass and phase_2_calc orif m lt 2 for task calc_biomass_and_LFs When sub population biomasses are calculated then any stratum in which p lt 0 3 is excluded but only for these biomasses see Section 5 3 1 41 5 1 4 User preferences for fish density variables When command preferences is used see Section 3 1 4 it specifies the user s preferences for the variables d w and C For each variable and station the option used will depend on what data are available The first option should be used unless the required data are not available in which case the second option is used unless the required data are not available etc All stations for which it is not possible to calculate all three variables d w C are dropped from the analysis and an appropriate error message is included in the main output file The options for each variable and the associated values are given in Table 6 Table 6 Three variable with user supplied calculation options the associated sub command of preferences and the value of the variable according to the option used Variable Sub Command Option Value d distance_towed recorded_distance distance in t_station from_lat_lon calculated from following attributes in t_station lat_s NorS_s long_s EorW_s lat_f NorS_f long_f EorW_f from_recorded_speed speed in t_station x t from_constant_sp
50. is restriction can be added as t_station gear_perf lt 3 and station_no not in 32 33 Similarly suppose you had completed 100 phase 1 stations stations 1 100 and were part way through phase 2 when you wanted to repeat the phase 2 calculations Then you could use subcommand t_station station_no lt 101 to exclude phase 2 stations from these calculations 7 REFERENCES Bull B Dunn A 2002 Catch at age user manual v1 06 2002 09 12 NIWA Internal report 114 23 p Unpublished report held in NIWA library Wellington Bull B Francis R LC C Dunn A McKenzie A Gilbert D J Smith M H Bian R 2008 CASAL C algorithmic stock assessment laboratory CASAL User Manual v2 20 2008 02 14 NIWA Technical Report 275 p Francis R LC C 1984 An adaptive strategy for stratified random trawl surveys New Zealand Journal of Marine and Freshwater Research 18 1 59 71 Francis R I C C 2006 Optimum allocation of stations to strata in trawl surveys New Zealand Fisheries Assessment Report 2006 23 50 p Mackay K A 2000 Database documentation trawl NIWA Internal report 73 48 p Unpblished report held in the NIWA library Wellington Latest electronic version available on NIWA intranet at http seaspray niwa co nz stockmon database 20documents Vignaux M 1994 Documentation of trawlsurvey analysis program MAF Fisheries Greta Point Internal Report 225 44 p Unpublished report held in the NIWA library Wellingto
51. iven in Table 1 For the station file T the only obligatory columns are station_no and stratum and trip_code if there is more than one trip Which of the other columns from this table are required depends on command preferences e g column dist_doors is not needed if preferences width_swept is constant_doorspread For the station catch file X the only obligatory columns are station_no stratum and a fish density with a label like XYZ_kg_km2 where XYZ is a user specified species code which need not be an official species code There can be densities for more than one species e g HOK_kg km2 HAK_kg km2 The only other optional column is trip_code a user provided code which need not be an official trip code which is required only if the user wishes to analyse more than one survey Note that vulnerabilities and vertical availabilities can not be entered as columns in this file When a station catch file is used it is assumed that the fish density columns are already adjusted for vulnerability and vertical availability 32 4 OUTPUT FILES Most output from SurvCalc goes to the main output file Section 4 1 which 1s named in the command line e g myf ile in the example command line at the beginning of Section 2 If requested additional output goes to one or more flat files Section 4 2 or a catch at age data file Section 4 3 The precision of many numbers in the output files as specified by command output_precision is
52. lt 0 5 are included in the calculation of sub population biomasses but a warning message is printed Where p 0 3 the sub population catch for station i is estimated as EX je ijk Wa Pi EN ijk Win p 11 for each station with LF data and SGI peda ADA 12 for stations without such data The remaining calculations of densities ad and ad biomasses b and b and s e s for ads bs and b use the same equations as for population biomass i e equations 1 3 and 5 9 except that each population variable is replaced by 1ts sub population analogue as defined in Table 4 5 4 Calculating LFs The equation for calculating SN the scaled LF for the ith tow depends on which option has been selected with command LF_scaling DE 3 Ni unscaled 1000 me p SN 212 ue i pS z wn Py nl a 2 10 Ni p f numbers_per_1000 numbers_per_tow 13 numbers_per_hour numbers_per_km2 or numbers_in_population For stations with zero catches SN 0 for all and k How the stratum and overall LFs are calculated depends on the task If it is calc_LF s these are simple i e unweighted sums or averages with no correction for stations or strata without LFs gt SN i for unscaled LFs i LF k me a SN x m otherwise SN 14 44 yA SN i for unscaled LFs s LF si E te SN m otherwise These equations differ from those in trawlsurvey where SN and SN are always calculated as
53. mmands as being vulnerability subcommands Specifies a vulnerability for the given species at all stations in the given trip Notes Each vulnerability command block applies to one species in one trip The command may be omitted when the vulnerability of the species is 1 at all stations in the trip Examples The following command block specifies that the vulnerability of HOK in trip tan0601 is 0 9 and 0 8 for stations 23 and 25 respectively and 1 for all other stations vulnerability tan0601_HOK default_value 1 other_stations 23 25 other_values 0 9 0 8 default_value Default value for vulnerability of the given species in the given trip Command vulnerability Type Positive number Default 1 Effects Defines the vulnerability of the given species in all stations in the given trip except for those in other_stations other_stations Stations at which the vulnerability differs from the default value Command vulnerability Conditions Must be existing stations Type Numeric vector Effects Specifies stations at which the vulnerability differs from the default value other_values Vulnerability values that differ from the default value Command vulnerability Type Positive numeric vector of same length as other_stations Effects Specifies the vulnerabilities for those stations in other_stations 16 vertical_availability Vertical availability of a species at each station in a trip Label Conditions Effects Notes Examples de
54. n 51 8 APPENDIX 1 COMMAND BLOCK FORMAT The main input file for SurvCalc uses a command block format similar to that used in CASAL Bull et al 2008 That is this file consists of a series of command blocks which may occur in any order Each command block either consists of a single command starting with the symbol and its arguments or a command starting with 2 and an optional label and one or more subcommands i e command arguments or command label subcommand arguments subcommand arguments Blank lines are ignored as is extra white space between arguments Comments beginning with are ignored If you want to remove a group of commands or subcommands using then comment out the whole block not just the first line Alternatively you can comment out an entire block by placing curly brackets around the text that you want to comment out Put in a as the first character on the line to start the comment block then end it with All lines including line breaks between and inclusive are ignored These should ideally be the first character on a line but if not then the entire line will be treated as part of the comment block Don t put extra white space before a character which must also be the first character on the line Make sure the file ends with a carriage return Commands and subcommands must consist of letters and or underscores and must not contain
55. n some stations to different strata new or existing A trip code must be in the vector trips It is a fatal error if there is any overlap between the stations affected by a change_strata command and a Oreassign_strata command for the same trip Defines any following subcommands as being reassign_strata subcommands for the specified trip Each station in stations is reassigned to the corresponding stratum in strata Omit the label if only one trip is being analysed i e if trips is of length 1 In the following example station 23 is reassigned to stratum 0012 and station 37 to stratum 012A reassign strata tan0601 stations 23 37 strata 0012 012A Numbers of those stations which are to be reassigned to different strata reassign_strata Each number in stations must be an existing station number for the specified trip Integer vector 14 strata Command Conditions Type new_strata Label Effects Notes Example strata Command Conditions Type Effects Notes areas Command Type Effects Notes change_stratum_area Label Effects Notes Example strata Command Conditions Type new_areas Command Type Names of the strata to which stations are to be reassigned reassign_strata Each string in st rata must be an existing stratum name for the specified trip or must be defined in command new_strata i e must be in new_strata trip names String vector Define new strata A trip code must be in the
56. n tables stratum_summary biomass sub_biomass_by_stratum biomass_by_species and biomass_by_species_stratum all scaled numbers in LF tables i e all columns whose labels start with SN all columns in biomass or LF tables whose labels start with cv all numbers in gain tables described in Section 4 1 1 38 5 CALCULATIONS IN SURVCALC This section contains detailed descriptions of the six types of calculation made by SurvCalc of densities Section 5 2 biomasses Section 5 3 LFs Section 5 4 phase 2 gain Section 5 5 projected c v s Section 5 6 and catch at age data Section 5 7 These are preceded by definitions of the data and notation used Section 5 1 5 1 Data and notation The data used in calculations come from three sources the main input file Section 3 1 flat files Section 3 2 and the trawl database see beginning of Section 3 1 1 The notation used in the calculations is explained in Tables 4 and 5 Table 4 Variables used in calculations their definitions and how they are calculated or where they come from Where a second variable is given in parentheses the first variable is for the whole population and the second is the analogous variable for a sub population See Section 5 7 for additional notation used for task output_for_catch_at_age Variable B b n m CF Definition overall biomass t number of strata in survey How calculated where from equation 7 number o
57. nd sub population biomass Ignored unless biomass is being calculated i e the task is calc_biomass or calc_biomass_and_LFs and more than on trip is being analysed Logical False A single table is output after all other outputs which combines the information in the biomass tables for each species trip combination Output table s of LFs by stratum output_tables Ignored unless LFs are being calculated i e the task is calc_LFs calc_biomass_and_LFs Logical False A separate table is output for each combination of species and trip Output table s of LFs by station output_tables Ignored unless LFs are being calculated i e the task is calc_LFs calc_biomass_and_LFs Logical False A separate table is output for each combination of species and trip Number_measured Output table s of numbers of fish measured by stratum and sex Command Conditions Type Default Notes LF_totals Command Conditions Type Default Notes output_tables Ignored unless LFs are being calculated i e the task is calc_LFs calc_biomass_and_LFs Logical False A separate table is output for each combination of species and trip Output table s of LF totals by stratum output_tables Ignored unless LFs are being calculated i e the task is calc_LFs calc_biomass_and_LFs Logical False A separate table is output for each combination of species and trip 26 or or or or output_precision Label Conditions
58. ngth weight coefficients independent of sex for HOK in trip tan0601 lw_coeff tan0601_HOK a 0 006 b 2 85 Length weight coefficients to calculate the weight of a fish in g from its length in cm lw_coeff Use either a b and optionally c or a_male b_male a_female b_female and optionally c_male c_female a_unsexed b_unsexed c_unsexed Numeric c 1 no default for a or b 19 a_male b_male c_male a_female b_female c_femaleLength weight coefficients by sex to Command Conditions Type Default calculate the weight of a fish in g from its length in cm lw_coeff Use either a b and optionally c or a_male b_male a_female b_female and optionally c_male c_female a_unsexed b_unsexed c_unsexed Numeric c_male 1andc_female 1 no default for other coefficients a_unsexed b_unsexed c_unsexed Length weight coefficients for unsexed to calculate the Command Conditions Type Default Notes weight of a fish in g from its length in cm lw_coeff Use either a b and optionally c or a_male b_male a_female b_female and optionally c_male c_female a_unsexed b_unsexed c_unsexed Numeric c_unsexed 1 no default for other coefficients If length weight coefficients are presented for males and females but not for unsexed fish then the weight of an unsexed fish of a given length is calculated as the average of the weights of a male and female of that length 2
59. on Number_measured Optional table produced for each species trip combination One line per stratum and a final summary line containing the number of fish measured by sex and stratum stratum Nfemale Nmale Nunsexed Nall 0001 624 656 1 1281 0002 302 131 0 433 All 2501 1765 9 4275 LF_totals Optional table produced for each species trip combination One line per stratum and a final summary line containing LF totals and their c v s by stratum and sex stratum SNfemale SNmale SNunsexed SNall cv_female cv_male cv_unsexed cv_all 0001 1163734 1 1463051 0 5269 5 2632054 6 39 44 100 35 0002 161545 6 69949 2 0 231494 9 17 8 0 36 All 7732192 2 5210113 8 39712 7 12982019 0 12 18 70 12 gains Obligatory table for task phase_2 calc produced for each species One line per stratum number of column determined by subcommand n_stations stratum 2 3 4 5 6 7 8 9 10 0001 16 8 T12 8 6 4 7 cre Bet 235 2 2 1 8 0002 8 5 4 3 2 6 Ly 5 a 0 9 0 7 0 6 O15 0 4 4 2 Output to flat files Flat files contain columns of data with the column names in the first line Each row of the flat file corresponds to a stratum with s or y or a station with t or w or a catch record with u or a subcatch record with w or a length record with v For s t u v or w the columns are the same as those extracted from the corresponding table in the database see Section 3 1 1 or as provided by the user with S T U V or W 4 2 1 Station
60. ons at which the area fished differs from the default value area_fished Must be existing stations for the given trip Numeric vector Specifies stations at which the area fished differs from the default value Area fished values that differ from the default value area_fished Positive numeric vector of same length as other_stations Specifies the areas fished for those stations in other_stations Areal_availability of a species at each stratum in a trip trip_code _ species_code For each combination of trip and species to be analysed there must not be both an areal_availability command block anda population_area command block use one or the other or neither Defines any following subcommands as being areal_availability subcommands Specifies an areal availability for the given species at all existing strata in the given trip Each areal_availability command block applies to one species in one trip The command may be omitted when the areal availability of the species is at all strata in the trip The following command block specifies that the areal availability of HOK in trip tan0601 is 0 8 and 1 2 for strata 0003 and 0004 respectively and 1 for all other strata areal_availability tan0601_HOK default_value 1 other_strata 0003 0004 other_values 0 8 1 2 Default value for areal availability of the given species in the given trip areal_availability Positive number 1 Defines the areal availability of the given speci
61. pecies No All stratum area area_fished trip amp species Yes 1w_coeff species or trip amp species No Depends on preferences Only needed for trip species combinations where length weight coefficients are required see Section 5 1 6 Use species label when the same coefficients are to be used for all trips 31 3 2 Other input files All input files other than the main one see Section 3 1 are flat files That is files containing columns of data with the column names in the first line Columns should be separated by white space i e spaces or tabs and the order of the columns doesn t matter Missing values should be entered as NULL The user specifies which flat files if any SurvCalc should read using the run time arguments S T U V W or X see Section 2 Each row of the flat file corresponds to a stratum with S or a station with T or X or a catch record i e for each station there must be a row for each species caught at that station with U or a subcatch record i e for each station there must be a row for each subcatch of species caught at that station with W or a length record i e for each station species and subcatch there must be a row for each length that occurs in the LF in that subcatch of that species at that station with V The columns and column names required for the stratum S catch U subcatch file W and length V files are precisely those g
62. r some lengths were measured or total catch was zero the n strata in the survey the strata in which some lengths were measured 5 all length records 7 and all sexes k for station i PB values of j and k such that for station i the length L and the sex k are within the required sub population values of j such that for station i the length L l i all lengths measured at station i EITHER all lengths for k 4 if a single set of length weight coefficients is provided 3 g MMM MMMMMMM OR all lengths for k 1 2 3 if length weight coefficients are provided by sex 40 5 1 1 Note on subcatches The notation Nix Lj and formulae see below used here for data from the table t_lgth are a bit more complicated than those of Vignaux 1994 because they allow for the possibility of what are called in database trawl subcatches which is to say more than one length sample for a given species from a given station e g one species in a catch may be divided roughly into two parts small fish and others with each part being weighed separately and having its own random length sample This means that there may be two or more records in t_lgth with the same values of trip_code station_no species and lgth The sum ae used in calculating an LF for each tow see below combines such records In trawlsurvey this summation is not done and this produces anomalies in the output station LFs as sho
63. rate data from a flat file rather than a database default value for infile is station_catch in output a flat file of stratum data one line per stratum default value for infile is stratum out output a flat file of station data one line per station default value for outfile is station out output a flat file of catch data one line per catch record default value for outfile is catch out output a flat file of length data one line per length record default value for outfile is lgth out w outfile output a flat file of subcatch data one line per subcatch record default value for outfile is subcatch out to be used only with o x outfile output a flat file of combined station catch and catch rate data one line per station default value for outfile is stat ion_catch out to be used only with b B or l y outfile output a flat file of combined stratum catch and catch rate data one line per stratum default value for outfile is stratum_catch out to be used only with b B or l f infile alternative name for the main input file e g ORH slc means that the main input file will be ORH sic rather than the default input s1c SurvCalc obtains the data to be analysed either from an Empress or PostgreSQL database or from flat files S T U V W and X but not both e g you cannot provide station and stratum flat files but expect SurvCalc to get catch and length data from the database The flat f
64. rrectly 1 3 Possible future extensions to SurvCalc This section describes features that may be incorporated in future versions of SurvCalc depending on demand from users and the coding effort required 1 SurvCalc could be extended to analyse survey data in databases like scallop and oyster The structures of these databases are broadly similar to trawl but with some relatively minor differences that would have to be allowed for 2 When designing new trawl surveys there is a need to decide on the basis of previous survey data how many stations need to be allocated to each stratum to achieve a target c v One way this is currently done is in the following two steps a extract data from previous surveys in the trawl database and b run the Splus function allocate see Appendix 2 It might be useful to combine these two steps in SurvCalc 3 SurvCalc allows the calculation of biomasses of sub populations defined by sex and length e g all males less than 30 cm long This could be extended to allow the use of gonad stage data in defining sub populations e g all females of stage gt 2 This would involve using trawl tables t_fish_bio and or t_lgth_stage not currently used by SurvCalc 4 SurvCalc could calculate length weight coefficients for a species in a survey or surveys Ideally this calculation should a be robust to outliers b include graphical output to show the user how well the data fit the estimated curve and
65. s_and_LFs Type Logical Default False Notes A separate table is output for each combination of species and trip but excluding species for which no sub populations have been defined biomass_by_species Output table s of all population biomass estimates from a trip by species and stratum Command output_tables Conditions Ignored unless biomass is being calculated i e the task is calc_biomass or calc_biomass_and_LFs Type Logical Default False Notes A separate table is output for each trip in which more than one species is analysed and this includes only those species in species trip codes This table does not include estimates of sub population biomass biomass_by_species_stratum Output a table of all population biomass estimates by species Command Conditions Type Default Notes and stratum output_tables Ignored unless biomass is being calculated i e the task is calc_biomass or calc_biomass_and_LFs Logical False A separate table is output for each trip in which more than one species is analysed and this includes only those species in species trip codes This table does not include estimates of sub population biomass 25 biomass_by_species_trip Command Conditions Type Default Notes LFs_by_stratum Command Conditions Type Default Notes LFs_by_station Command Conditions Type Default Notes estimates by species and trip output_tables Output a table of all population a
66. see what tables are there simply type names hokout buc8501 which may return something like 1 SELECTIONS RUN PARAMETERS ORH OEO hokout buc8501SSELECTIONS would contain information about the strata and stations selected for this trip and hokout buc8501 RUN PARAMETERS would contain the user supplied preferences for distance towed width swept etc Similarly typing names hokout buc8501 0RH lists the output tables for ORH in trip buc8501 For a description of all tables that can occur in a main output file from SurvCalc see Section 4 1 allocate Function allocate is useful at the survey design phase It uses historical data from previous surveys in the same area to work out how many stations should be allocated to each survey stratum It can be used in two different ways as illustrated by the following calls allocate constraint cv limit 20 strata mystrat data mydat allocate constraint stations limit 20 strata mystrat data mydat The first call calculates the minimum number of stations and their allocation to strata to achieve a c v of 20 whereas the second calculates the best allocation given that there will be a total of 20 stations In both cases myst rat should be a dataframe containing stratum information in the format stratum area minimum maximum 0001 1245 3 20 0003 347 3 15 53 with the last two columns specify the minimum and maximum
67. sensotesnetieeeuetsuaneeyedess 10 3 1 Phe main input lena ias dd rl eiii 10 3 1 1 Commands defining the data oooncnnnncnnnnnnncnnoncnoncnoncnonacononononnnnnnnn crac cra nrnn cra 11 3 1 2 Commands modifying the data ooooonccnncnincnnoncnoncnoncnonaconacanononccnnnnnnn crac conan 14 3 1 3 Commands extending the data oonconccnnncnincnnnccconcnoncconaconnnoncconcnnnnanan crac cono cra 16 3 1 4 Commands defining the calculatiOMS oonnonnncninnnnncononnnonnconnconnnnnnnnnc crac crac crac 21 3 1 5 Commands defining output ocoooccnnccnncononnnonnnonnconncno nono nocnonnnn cnn nannnnnnc cra nica cra 25 3 1 6 Examples of main input files oonnonnnnnnncnoccnoncnonanonanonononononnnnnn crac nr nacrna conan 27 3 1 7 Repeated commands in the main input file oooooonnonncnncnncnnncnnncnnacnnacnnanccnnno 31 32 Other input file Sranie is E A aiii 32 OUTPUT HLES ccncoriontsicnaiaicminarondodda nd cr neo iria a dra E deine ci 33 4 1 Marin output erstes 33 4 1 1 Tables in the main output file eee ceeceseceeceeceeeeseeseeesseeseneeeneeeaaeenaes 34 4 2 Outputto Gat tiles cuina acesteidelenstaidielaeselveeslsahvnsilebnsiidlecbatelpeaie 36 4 2 1 StatiOm Catch file sy sssisessdess cessuesscesdesssnsveed ssedech ce aeeuvidadesusdesasedecdsaetudeseasouvacanesbees 36 4 2 2 Output to stratum catch file eee ceeeceseceseceeeeeseeeeeeeeaeeeaeeeaaecaecnaeeeaeeeaeee 37 4 3 Catch at age data Outputs i cccsssitatissscapesasentaeen tedesaand Eas
68. st the survey strata Some guidance on how SurvCalc should be used during a 2 phase survey is given in Section 6 4 Task phase_2_calc Calculates separately for each species requested the relative gains in terms of reduced variance of biomass estimates associated with allocating varying numbers of phase 2 stations in each stratum From this information the optimum phase 2 allocation can be derived for each species The last three tasks simply reorganise the survey data and output it in a different form 5 Task output_flat_files Output data in one or more of seven types of flat file Depending on the type each line of an output flat file may represent a stratum a station a catch or subcatch record 1 e a combination of a station and a species or a length record i e a combination of station species subcatch and length 6 Task output_LW_coeffs Output a table of length weight coefficients This shows what length weight coefficients are held in database rdb for each species so that the user can decide whether to use these stored coefficients or to specify new coefficients See Section 5 1 6 for a description of how these coefficients are used in various calculations 7 Task output_for_catch_at_age Output a file in either survey or survey sub format whichever is appropriate for input to the catch at age software Bull amp Dunn 2002 That is a file that can be read by the catch at age function import length dat
69. sums for a calc_LFss analysis That is SV up Nin and SN 2 ip SN i SN 15 When the task is calc_biomass_and_LfFs the stratum LFs are calculated by averaging the station LFs scaling using the population area and then adjusting for strata with no LF data SN y MSN K 16 where MSN AN m 17 AD panc 18 i AD p and AD s Done AD m 19 They are then corrected so that the estimated numbers of fish are consistent with the estimated biomass and length weight relationships SN mi SN 4 CF 20 where CF 10 B Y SN Q1 Overall LFs are calculated using B SN s LF SN y 22 2 s LF B note that for surveys in which there are LF data in all strata this equation becomes a simple sum SN X SN and then corrected for consistency with the estimated biomass and length weight relationships SN SNCF 23 where CF 10 B Y SN W 24 5 4 1 Calculating c v s for LFs C v s for LFs are calculated only when the task is calc_biomass_and_LFs and then only for corrected LF totals TN for each stratum and TN for the whole population using the equations v TN 100s e MTN MTN 25 45 100 e MTN KY and c v TN oo e a s k 26 PA MTN K 2 0 5 Dio Ne MN gs where s e MTN m m 1 27 The derivation of equations 26 27 is worth explaining These c v s take into account only the uncertainty arising from the var
70. tabase Command Conditions Type Default Effects hostname Command Conditions Type Effects schema Command Conditions Type Default Effects database_name Command Type Default Effects t_station_columns Command Type Conditions Effects Defines any following subcommands as being input_from_database Ignored if the user has provided flat file input data with run time arguments S T U V or W database input_from_database Either Empress or Postgresql String Empress Specifies the database from which the data are to be extracted Postgresql server name input_from_database Only used for Postgresql database and only needed when this database is being accessed across a network i e when SurvCalc is not running on the machine on which the Postgresql database resides String Specifies the machine name on which the Postgresql database resides schema name input_from_database Only used for Postgresql database String trawl Specifies the Postgresql schema name under which data tables are stored database name input_from_database String trawl for Empress or fish for Postgresql Specifies the database name in which data tables are stored the additional attributes to be extracted from t_station table input_from_database String vector The attributes must exist in t_station table of the database Specifies the additional attributes to be extracted from t_station table 12 where Lab
71. the catch at age software can be output The analysis of potting surveys is more straightforward and sensible i e the user will not have to make up fake values for doorspread and distance towed Sex specific length weight coefficients are allowed The calculations for phase 2 of a 2 phase survey are much more extensive see Section 6 The user can control the degree of precision expressed as the number of significant figures and or decimal places of each type of output 1 2 2 Excluded features of trawlsurvey 1 trawlsurvey produces LFs as percentages in the main output file and as numbers in separate files but some summary information for the numbers LFs is confusingly included in the main output file SurvCalc produces LFs only as numbers LFs as percentages are easily calculated from these 2 SurvCalc does not allow the user to define bounds and interval for LFs e g for lengths 20 cm to 50 cm in 2 cm steps All SurvCalc LFs cover the full range of the data in 1 cm steps The length bounds are not well handled in trawlsurvey it is not made clear that oddly they apply to the percentage LFs and confusingly the summaries of the numbers LFs but not to the numbers LFs also the user is not informed when there are length data outside the specified length bounds 3 trawlsurvey outputs a table containing inter alia mean fish densities and biomass estimates by stratum where the density units kg km and k
72. ting strata can be changed change_strata Label Conditions Effects Notes Examples from Command Conditions Type to Command Conditions Type reassign_strata Label Conditions Effects Notes Examples stations Command Conditions Type Reassign all stations in some strata to different strata new or existing A trip code must be in the vector trips It is a fatal error if there is any overlap between the stations affected by a change_strata command and a reassign_strata command for the same trip Defines any following subcommands as being change_strata subcommands for the specified trip All stations from the specified trip that were originally assigned to one of the strata listed in to are reassigned to the corresponding stratum in from Omit the label if only one trip is being analysed i e if trips is of length 1 The following example assigns all stations in stratum 0023 or 0025 to stratum 023A and all stations in stratum 0027 to stratum 0030 change_strata from 0023 0025 0027 to 023A 023A 0030 Names of strata whose stations are to be reassigned change_strata Each string in from must be an existing stratum name for the specified trip String vector Names of strata to which stations are to be reassigned change_strata Each string in must be either an existing stratum or defined in command new_strata 1 e must be in new_strata trip names String vector of same length as from Reassig
73. tion catch data W This command must be provided for all trips in which constant_speed time occurs in preferences trip distance_towed It will be ignored for other trips Defines any following subcommands as being constant_speed subcommands for the specified trip Omit the label if only one trip is being analysed e if trips is of length 1 User supplied vessel speed for the given trip constant_speed Number This vessel speed will be used only for those stations in the specified trip in which constant_speed time is the selected option for calculating distance towed as specified in command preferences 22 constant_doorspread Doorspread to be used in calculating the width swept Label Conditions Effects Notes value Command Type Notes ELF_scaling Conditions Type Default Effects Notes phase_2 Label Conditions Effects Notes algorithm Command Type Default Effects Notes n_stations Command Type Default Effects A trip code must be in the vector trips Ignored if the user has provided station catch data W This command must be provided for all trips in which constant_doorspread occurs in preferences trip width_swept It will be ignored for other trips Defines any following subcommands as being constant_doorspread subcommands for the specified trip Omit the label if only one trip is being analysed i e if trips is of length 1 User supplied vessel doorspread for the given
74. tratum area for all strata in the trip The following command block specifies that the population area of HOK in trip tan0601 is 2957 and 1325 for strata 0003 and 0004 respectively and equal to the stratum area for all other strata population_area other_strata 0003 0004 other_values 2957 1325 Strata in which the population area differs from the stratum area population_area Must be existing strata for the given trip String Specifies strata in which the population area differs from the stratum area Population areas that differ from stratum area population_area Positive numeric vector of same length as other_strata Specifies the population area for those strata in other_strata Length weight coefficients for a species in a trip trip_code _ species_code Needed only if length weight coefficients are required for an analysis and the user wants to use values different from those in database rdb If this command is used it must be repeated for each combination of trip and species for which length weight coefficients are required see Section 5 1 6 If it is not used then all required length weight coefficients will be read from rdb Defines any following subcommands as being Q1w_coeff subcommands Specifies the coefficients used to calculate the weight of a fish in g from its length in cm Note that length weight coefficients are not always required in analyses see Section 5 1 6 The following command block specifies the le
75. trip constant_doorspread Number This doorspread will be used only for those stations in the specified trip in which constant_doorspread is the selected option for calculating width swept as specified in command preferences Method of scaling LFs Ignored unless LFs are being calculated i e the task is calc_biomass_and_LFs or calc_LFs Option numbers_in_population is obligatory if the task is calc_biomass_and_LFs and not allowed if itis calc_LFs String must be one of unscaled numbers_per_1000 numbers_per_tow numbers_per_hour numbers_per_km2 numbers_in_population numbers_in_population in B run no defaultin 1 run Defines how LFs are scaled see Section 5 4 This command applies to all trips and species analysed It can be omitted if the task is calc_biomass in which case there is only one LF_scaling option User preferences for phase 2 gain calculations A species code must be in species trip code Ignored unless the task is phase_2_calc Defines any following subcommands as being phase_2 subcommands If this command is used after some phase 2 stations have already been carried out then you should use command where to exclude these stations from this analysis It is not necessary to specify a trip for the phase 2 calculations because it is a fatal error to request phase_2_ calc when trips is of length gt 1 Algorithm to be used in calculating gains phase_2 String must be either mean_squaredor
76. understand that allocate is a very simple function which makes no allowance for the very real possibility that the distribution of fish in the survey being planned may be quite different from that in the historical surveys For this reason it is prudent to be conservative in its use For example it is common to use allocate to determine a station allocation that will achieve the target c v s in phase 1 of a 2 phase survey Then if fish distributions during the survey differ substantially from those in previous surveys so that the target c v s are not met during phase 1 there is still a possibility that they will be met by the end of phase 2 54
77. used in the calculation of the usual biomass c v s 1 e equations 8 and 9 except that for strata that are not yet completed i e those specified in the projected_cvs command block s e AD or s e ad in the case of sub populations in equation 8 is replaced by S prj ADs or S proj ads where these are defined by o AD s e AD n n e 32 and 8 ad s e ad n n e Js 33 47 5 7 Output for catch at age With task output_for_catch_at_age SurvCalc must read additional material from the database beyond what is given in Table 1 and may have to do some additional calculations The first additional material that must be read from the database is column subcatch_no from database table lgth If this column is always equal to 1 there are no subcatches so the output will be in format survey and no additional calculations are required The column kg km2 in the output file see Section 4 3 will contain AD If there are any subcatches the output will be in format survey sub column subsample in the output file will be subcatch_no from table lgth and column subsample_kg km2 will contain the subcatch catch rate AD where c indexes the subcatches for the species in question at station i and AD 10 C f Where Cie comes from depends on what option is being used for catch_weight as specified in the Qpreferences command see Section 3 1 4 With option recorded
78. what range of lengths is well covered by the relationship and c include the ability to test for significant differences between the parameters for males and females 5 Some users have asked for the ability to calculate total biomass for large groups of species e g all fish which means excluding invertebrates etc This might involve using the attribute descrptn in table species_master in database rdb to define groups of species 6 Current options for LF scaling make no allowance for correlation in the samples i e the fact that typically the lengths of two fish from the same tow are more similar than those from different tows When more sophisticated scaling schemes are developed they should be available in SurvCalc 2 RUNNING SURVCALC SurvCalc is run from the command prompt in Windows or Unix by typing a command like SurvCalc b t stnfile gt myfile It uses information from the main input file and possibly some flat files all described in Section 3 makes certain calculations documented in Section 5 and writes results to the main output file myfile in the preceding example and possibly other files see Section 4 In the command line between SurvCalc and gt there must be one or more run time arguments which may occur in any order as described below The command line must include exactly one of following run time arguments which describes the task required of SurvCalc Argument Task Task description
79. wn in the following example in which there are two lines for fish of length 42 cm in station 79 However this error affects only the station LFs the stratum and overall LFs are OK stn str L Nall Nmale Nfemale Nunsexed SNall SNmale SNfemale SNunsexed 79 0017 41 T2 8 4 0 338 11 225 41 112 70 0 79 0017 42 9 3 6 0 253 58 84 53 169 06 0 79 0017 42 l 1 0 O 14 77 14 77 0 00 0 79 0017 43 11 8 3 0 309 94 225 41 84 53 0 5 1 2 Note on stations and strata without LF data There can be two types of stations or strata without LF data A those with no catch of the target species and B those with a non zero catch from which no fish were measured It is important to notice that both types are excluded from the summations gt for stations and a for strata but that only type B is excluded in defining the variables m and m and in the summation a eee 4 l 5 1 3 Excluding stations and strata For tasks calc_biomass calc_LFs calc_biomass_and_LFs and phase_2_calc SurvCalc sometimes excludes some stations and or strata because of inadequate data A station is excluded if A it is not possible to calculate the fish density AD because of missing data e g distance towed or width of tow missing with no default provided or B the stratum it is in is excluded Note that criterion A does not apply when the task is calc_LFs and LF_scaling is not numbers_per_km2 because fish densities are not calculated in this case The criter

SurvCalc manual 12March2012 - These are not the droids you are

Contents

Download Pdf Manuals

Related Search

Related Contents