ICES CM 19971H:5
Pelagic Fish Committee
.
REPORT OF TUE
PLANNING GROUP FOR HERRING SURVEYS
Aberdeen, United Kingdom
24-28 February 1997
This report is not to be quoted without prior consultation with the
General Secretary. The document is areport of an expert group
under the auspices of the International Council for the Exploration of
the Sea and does not necessarily represent the views of the Council.
International Council for the Exploration of the Sea
Conseil International pour l'Exploration de Ia Mer
Pala:gade 2-4
DK-1261 Copenhagen K
Denmark
TABLE OF CONTENTS
Section
Page
INTRODUCTION
1.1 Terms of reference
1.2 Participants
1.3 An outline of the problem in the assessment.
.
,),
"
2 REVIEW OF THE SURVEY TIME SERIES
2.1 Results of the studies
2.1.1 The review of amplitude distributions from Orkney-Shetland area
2.1.2 The distribution of abundance from acoustic surveys
2.1.3 Comparison between acoustic survey and IBTS time series
2.1.4 Missing catch model
1
2
2
2
2
2
3 ,USE OF HERRING ACOUSTIC SURVEY IN ASSESSMENT
3.1 Remaining unanswered questions
3.2 Conclusions from the studies
3
3
3
4 ADVICE AND FUTURE WORK TO RESOLVE THE PROBLEMS
3
5
•
1
1
1
1
REVIEW OF LARVAE SURVEYS
_
4
6 COORDINATIONOFLARVAESURVEYS
6.1 Surveys planned for 1997/98
6.2 Requirements for desired complete coverage in 199912000
5
5
5
7 FUTURE DATA PROCESSING NEEDS FOR THE LARVAE SURVEYS
5
8 COORDINATION OF ACOUSTIC SURVEYS
6
9 FUTURE DATA PROCESSING NEEDS FORACOUSTIC SURVEYS
9.1 Herring abundance data
9.2 The workshop on scrutinising echograms
9.3 Intercalibration
9.3.1 Procedure far the intercalibration of echosounders during the Narth Sea herring survey
9.4 Exchange of length and age data from trawl hauls
6
6
6
7
7
8
10 RECOMMENDATIONS
8
11 REFERENCES
9
12 APPENDICES
11
APPENDIX A - Amplitude distributions far Scotia surveys in IVa 1987-1996
12
APPENDIX B - Abundance, and Biomass of herring by ICES statistical rectangle from acoustic surveys
1984 to 1996
25
APPENDIX C - Comparison of acoustic and IBTS times series by examination of relative cohort strength........ 52
APPENDIX D - Perceptions of North Sea herring stock dynamics and survey variability that are robust to
catch misreporting
57
APPENDIX E - Effects of reduced sampling effort on abundance and production estimates from
North Sea Herring Larvae Surveys
E:\PFaPGHERS97\REP.DOC
01/04/97
70
Section
Page
APPENDIX F - Format for exchange of trawl sampIe herring age and length dutu
73
APPENDIX G - Communicution information for research vessels
79
APPENDIX H - Planning Group for Herring Survey contact numbers
80
•
E:\PFaPGHERS97\REP.DOC
01104/97
ii
1
INTRODUCTION
1.1
Terms ofreference
The planning group for herring survcys will meet in Aberdcen, UK, from 24 to 28 Fcbruary 1997 to:
a) Coordinate the timing and area allocation of, and methodologics for, acoustic and larvae surveys for herring
in the North Sea Divisions VIa and lIla, and the Westcrn Baltic;
b) Combine the survey data to provide estimates of abundance for the populations within the area;
c) Evaluate the usefulnes~ ofthe herring acoustic time scrics with respect to North Sea assessment;
d) Discuss the outcome of studies of the consequences of reduccd effort and area covcrage for the herring larvae
surveys;
e) Define future data proccssing nceds for combining proposed acoustic and larvae surveys data from different
countries and whcre this should be carricd out ovcr the next few years;
•
f) Develop a proposal for a survcy plan for acoustic and larvae surveys which wiII provide data required far
future North Sea assessments.
1.2
Martin Bailey
Bram Couperus
Paul Femandes
Eberhard Götze
Nils Häkansson
Cornelius Hammer
Kenneth Patterson
David Rcid
John Simmonds (Chairman)
Karl-Johan Strehr
Reidar Toresen
Else Torstensen
1.3
•
Participants
United Kingdom
Netherlands
United Kingdom
Gcrmany
Swedcn
Gcrmany
Unitcd Kingdom
United Kingdom
Unitcd Kingdom
Denmark
Norway
Norway
An outline of the problem in the assessment
Norlh Sca herring stock assessmcnts from 1990 onwards show a systcmatic ovcrcstimation of the spawning stock
biomass (Anon, 1996a). During the years 1990 to 1995, the spawning stock biomass estimatcs (and consequent
catch forecasts) have bcen rcviscd successively downwards. The reasons for this were thought to be associated at
least in part with anomalously low acoustic survey obscrvations in 1987 and 1988, followed by relatively higher
observations in the period 1989 to 1995. Revisions in the assessments are shown in Figure I, which also shows
the trend in acoustic survcy stock size estimates for comparative purposes. After 1989, the acoustic survey
biomass estimatcs are much highcr than the assessment working group's population model cstimatcs, whcreas
before 1988 the cstimatcs are rather similar.
The assessment working group identified an increase in acoustic survey efficiency, and possible misreporting of
catches as plausible factors as probable cause for this ovcrestimation. In consequence, the assessmcnt working
group rccommended improved provision of information on catches and on survey cstimates of stock size.
2
REVIEW OF THE SURVEY TIME SERIES
Four studies were prescnted:
•
A review of the amplitude distributions [rom the acoustic surveys in the Orkney-Shetland area from 1988 10
1996. The review is documented as Appendix A to the report;
E:\PFarGHERS97\REP.DOC
01104/97
•
•
•
A review of the spatial distribution of abundance for the fu1l sequence of acoustic surveys from 1984 to 1996.
The data from a1l surveys has been entered as numbers and biomass at age and maturity by leES statistical
rectangle and is available as aseries of Excel spreadsheets. The spawning stock abundance and biomass are
documented in Appendix B to the report;
A review of the acoustic survey time series age disaggregated index with reference to the IBTS age
disaggregated index. This review is inc1uded as Appendix C to the report;
A missing catch stock model was presented, this is inc1uded as Appendix D to the report.
2.1
Rcsults of the studies
2.1.1
The review of amplitude distributions from Orkney-Shetland area
A number of conc1usions were presented:
I. The ratio of the number of zero and minimum dass values changed through the period of study; the number
ofzero values increased.
2. The skew factor for the distribution increased during the period of the study.
3. The number of zero rectangles was greater after 1990.
Items land 3 are incompatible with an increase in abundance due to changes in data treatment or due to changes
in the mean as an cstimator of the stock abundance value. However, there is a possibility that item 2 may be
caused by underestimation of the largest schools in the early years duc to saturation of the highest signals in the
clectronics, this could cxplain a changc in survcy cfficiency bctwcen 1990 and 1991.
2.1.2
•
The distribution ofabundanee from aeoustic survc)'s
Thc distribution maps show importunt changes in distribution both across the North Sea and east and west of
Shetland. The maps show that the survey in 1988 had substantial high values on the northern boundary and this
may havc given rise to a low estimatc in this year due to a lack of coverage.
The distribution shows some year (0 year variation in the abundance in the area west of Orkney- Shetland and
north of the Minch. There is uncertainty as to the correct a1l0cution of these fish to the North Sea or west of
Scotland stocks.
2.1.3
Comparison bctwccn aeoustic suncy and IßTS time serics
The ratio of the acoustic index with the IBTS from 1987 to 1994 shows considerable fluctuation with a low point
in 1988, resulting in a factor of 1.7 or 1.2 between observations at thc ends of this period. The difference
depends on the method uscd to combine the year classes. The differences over the fu1l available time serics from
1984 to 1994 indicatcs a factor bctwccn 1.4 to 0.7 from the mid 1980s to the early 1990s. The study also
examined the prccision for the estimates of year-dass strength, these are not of high quality but do suggest that
therc is considerable overlap in the series and the acoustic series provides a more precise estimate of year-c1ass
strength at 2 to 4 ring.
2.1.4
l\lissing eateh model
A population model similar in structure to the working group's assessment model, but exduding catch
information, was used to investigate whether the perceptions of increasing catchability in the acoustic survey
biomass estimate are dcpcndent on using reported catches in a VPA-type model structure. Some estimates ofthe
variability in different datu series were calculated. Detailed methodology and results are reported in Appendix D.
The f01l0wing inferences were drawn from the model fits:
1. The perception of increasing catchability with time for the acoustic survey biomass cstimates (with respect to
larvae surveys, to the IBTS index, or cven to the acoustic survey age-structure alone) remained, even when
reportcd catches, though not catch age structure, are excluded from the stock assessment model.
2. In terms of measures of variability in abundancc cstimates, thc IBTS abundancc index performs best, the
acoustic survey abundancc index performs worst, and the performance of the MLAI index is intermediate.
E:\PFC\PGHERS97\REP.DOC
01/04/97
2
•
- - - - - - -
•
3. In terms of measures of between-year correlation in errors in abundance, which may be more important in
terms of providing advice for management purposes, the IBTS survey is unlikely to have error correlation
(P=0.97), the acoustic survey is very likely to have correlated errors (P=0.02) and the MLAI index is
somewhat less Iikc1y to have correlated errors (P = 0.06).
4. In terms of estimating the age-structure of either the eatch or the stock, the acoustic survey performs best, it
has the largest effective sampie size (smallest t2), the IBTS performs worst, it has the smallest effective
sampies size, and the sampling of commercial catches is intermediate.
Overall the model suggests that the most reliable sourees of information are the aeoustic survey estimates of agestructure and the IBTS spawning biomass estimates. These inferenees are of course predicated on the
assumptions detailed in Appendix D (Section 2.1), and rely on ignoring process errors (eg changes in selection
pattern, ehanges in natural mortality, etc).
lt
•
3
USE OF HERRING ACOUSTIC SURVEY IN ASSESSMENT
3.1
Remaining unanswercd questions
a) Why is the age structure from the acoustic survey the most precise age index while the abundance index is the
most divergent, when the abundance estimates are used to derive the age structure for a stock with spatially
variable age structure?
b) Why does the IBTS abundance index perform best, during aperiod with changing adult age structure, when it
is dominated by a single year c1ass because it is derived from a survey with a fishing gear with a steep age
seleetion funetion?
c) Why does the acoustic abundance index whieh shows the least year to year fluctuation give a stock trajectory
that is different from other indices?
3.2
Conclusions from the studies
a) The problem of divergent indices is still present when the effect of the magnitude of unreported eatch, with a
linear increasing fishing mortality, is inc1uded in the analysis.
b) The acoustic survey and the IBTS survey indices may be more sclf consistent than all the indices combined.
•
c) There was a general increase in the frequency of zero values (2.5 Nm sampie values) in the acoustic survey of
the Orkney-Shetland area during the period 1987 to 1995. This would indicate a tendency to underestimate
the population. The increase in skew in the amplitude distributions during this period could be caused by
signal saturation for large schools, and thus eould explain underestimation during this period.
4
ADVICE AND FUTURE WORK TO RESOLVE TUE PROBLEMS
a) There is a need to investigate the importance in the survey time series of abundance changes to the west of
Orkney-Shetland and north of the Minch. If these are important the age and length structure of herring should
be investigated and these should be used to advise on the split between North Sea and west coast herring
b) An examination of the depth distribution of herring over the survey period should be carried out. These
should be investigated in the light of the possible depth dependance of herring target strength, to estimate
possible abundance changes over the survey period.
c) The use of General Additive Models (GAMs) on age disaggregated spatial distributions of herring from
acoustie and IBTS surveys should be examined to see if these can be helpfuI.
d) Inferences drawn from the age strueture and abundance indices may differ. This requires eare when the
indices are uscu in the assessment.
E:\PFaPGIIERS97\REP.DOC
19/03/97
3
..
~
.
i','
e) Perceptions of series divergence are dependant on the years, age ranges, and year dass weighting given to
different year dasses.
f) There is a need to carry out studies of the implications of saturation in the electronics on surveys prior to
1991.
g) There is a need to increase conlidence in the compatibility of multiple surveys used in the North Sea, Western
Baltic and VIa. For this purpose it is proposed to inc1ude intercalibration during the survey, to exchange data
on length and age distributions from hauls carried out during one year (1995) and to hold a workshop to study
the interpretation stage of acoustic survey echo sounder output allocation to herring.
5
REVIEW OF LARVAE SURVEYS
The substantial dec1ine in ship time and sampling effort allocated to the herring larvae surveys in recent years
required a study of the effects on the estimates of larvae abundance and production derived from these surveys.
The first step of this analysis was presented, considering a reduction in the number of subareas to be samplcd and
the required frequency of intermediate complete surveys (see Appendix E).
From the presentation and discussion of this study and comparison vlith results from a multiplicative model for
the abundance index LAI, the following main conc1usions can be drawn:
I. There is no long term stability in the relative importance of the different spawning areas and therefore the
assumptions required for the multiplicative model used to overcome the problem of missing values in the data
sets are not valid when based on extended time pedods. The inc1usion of interaction terms between survey
areas may alleviate this problem.
•
2. For the ca1culation of abundance indices it would be prudent to concentrate effort on a few target areas rather
than attempting to cover all spawning areas of the North Sea as has been done in thc past. The precision of
stock sizc estimates is not reduced when based on combined sampling results from Orkney-Shetland and
Buchan or southern North Sea as compared to inc1uding all three areas or a complete coverage.
3. Complete coverage would nevertheless be required though less frequently, to observe long term trends in the
relative importance ofthe different spawning areas and in the zIk values. From the multiplicative model there
is evidence for temporal periodicity in the residuals of thc larvae abundancc values of thc order of
approximately 6-8 years. In order to study this periodicity, complete coverage would bc required every three
years.
4. The residuals in the multiplicative model for thc abundance index (LAI) indicate that thc results from
different time periods within areas show differences similar to those between areas. It is thus not to bc
expected that a reduction in thc survcy frequency can be achieved without loss in precision of stock size
estimates based on the LAI. For LPE one coverage may bc suflicient, as has previously been suggested by
the Herring Larvae Survey Working Group (Anon, 1990). This has to be reviewcd, however, in the light of
an additional reduction in thc areas covered.
For thc larvae surveys the Planning Group recommends:
1. Yearly surveys should focus on the southern North Sea as well as on the Orkney-Shetland and/or Buchan
area. A more detailed analysis of the historieal databasc is required to elucidatc which of the two northern
areas should receive a higher priority.
2. Efforts should be made to organise for a complete coverage every three years, out of phase with the mackerel
egg survey, starting in 1999.
3. The effcct of survcy timing on larvac abundance indices ami production estimates should be examined in
more detail from the historieal database, to confirm or disprove the indications so rar available.
E:\PFOPGIIERS97\REP.DOC
19/03/97
4
•
4. Reliability and changes of the zIk values should be studied as the LPE is especially sensitive to this
parameter. A standard procedure to estimate zIk should be defined and the existing data series revised
aeeordingly.
.
•
6
COORDINAnON OF LARVAE SURVEYS
6.1
Surveys planned for 1997/98
Gennany
16-30 Sep 97
Orkney-Shetland
Netherlands
16-30 Sep 97
Buchan
Netherlands
01-15 Dec 97
Southem North Sea
Germany
01-15 Jan 98
Southern North Sea
Netherlands
16-31 Jan 98
Southern North Sea
6.2
Requirements for desired complete coverage in 1999/2000
Area
Period
Stations
Time (days)
Orkney-Shetland
01-15 Sep
110
*12
16-30 Sep
110
12
01-15 Sep
80
*7
16-30 Sep
80
*7
01-15 Sep
70
*6
16-30 Sep
75
*6
01-150et
110
*10
16-310et
110
*10
16-31Dec
60
5
01-15Jan
90
8
16-31 Jan
90
8
Buchan
Central North Sea
Southem North Sea
Optimal complete eoverage for calculating LAI and LPE would require a total of about 90 days survey time. The
survey time required in addition to that presently available is indieated in the above table by * and amounts to
about 58 days.
7
FUTURE DATA PROCESSING NEEDS FOR THE LARVAE SURVEYS
A copy of the herring larvae database has been sueeessfully transferred and implemented in Rostoek (Germany).
An implementation in Kiel (Germany) is intended as soon as all required information has beeome available for
rebuilding and eheeking the programmes for routine analyses of results from the yearly surveys. It is expeeted
that the routine analysis and reporting can be provided from Kiel from 1999 onwards. At the 1997 ICES meeting
it wiII be discussed ami deeided whether this task can be taken up at Kiel for the 1998/99 period.
E:\PFC\PGHERS97\REP.DOC
19/03197
5
8
COORDINATION OF ACOUSTIC SURVEYS
In 1997 the following surveys will be carried out in the North Sea and west of Scotland
Charter
12 - 29 July
1N0rth of 56°30'N west of 3°W
Dana
2-12July
North of 57° east of 6°E
GO Sars
28 June - 18 July
North of 5]0 east of I°W with reduced effort east of 3°E
Scotia
8 - 28 July
North of 58°30' between 4°W and 2°E
fridens
30 June - 18 July
South of 59°N west of 2°E
IW Herwig
23 June - 16 July
South of 57°N east of 2°E reduced effort between 2°_6°E
The following survey will be carried out in the western Baltic.
Solea
9
12 Sept - 20ct
ICES Sub-divisions 22, 23 and 24
FUTURE DATA PROCESSING NEEDS FOR ACOUSTIC SURVEVS
There are a number of developments requiring data processing, the need to deal more correctly with the herring
abundance data, the need for a workshop on herring echogram scrutiny procedures and the need to exchange
hcrring survey trawl data.
9.1
•
Herring abundance data
There is a necd to reorganise the data collation in order to obtain bctter distribution maps and better overall
combination of data. For this purpose, the planning group is organising the preparation of a herring survey
database under an EU project ECHOHER.
For 1997, data on number and biomass of herring by ICES statistical rectangle and age/maturity proportion will
be sent to John Simmonds in Aberdeen, Scotland. A blank Excel file will be provided. Data on Sprat will be
sent to Else Torstensen in Arendal, Norway.
9.2
The workshop on scrutinising echograms
In undertaking a herring acoustic survey each country covers aseparate area each with its specific characteristics,
such as spatial distribution and bottom conditions. When analysing the data the scientist allocates acoustic
signals to species (scrutinising). Thus decisions are made, based on experience gained by individual scientists
during surveys in specific areas. This indicates a subjective input to the analysis process.
In order to improve data analysis, a synchronisation of the way data are interpreted is required. The planning
group therefore recommends that a workshop on scrutinising be organised. It was suggested that this workshop
should be held in Bergen, January 1998, during the next planning group for herring surveys meeting.
At this workshop every country participating in the international herring acoustic survey should bring national
data for analysis. The data has to be the following:
•
•
•
•
•
a data set, typical for the area, containing one survey day and an optional 12 hours of difficult problems;
the paper output from the echosounder;
the BI 500 files on tape (a scrutinised version and a blank version). These should be sent to IMR Bergen in
August (8 rnrn Exabyte or QIC-150 format) for testing and control (contact Hans Peter Knudsen and Kaare A
Hansen);
the trawl data with the species composition (% in weight)*;
weather conditions and notes on circumstances that may be relevant to the data.
E:\PFaPGIIERS97\REPDOC
01/04/97
6
•
*In areas where thc bulk of thc SA-values comc from schools, thc interpretation of the net-sounder traces in
combination with actual trawl data are important, because thc composition of the trawl may be different from the
school composition. Herc thc interprctations of observations during thc tow are subjective. Thcreforc dctailcd
notes on trawl performance arc required.
The different data sets will bc analysed by a group of scientists from all countries involved and the results
compared to provide a measure of the precision of the serutinising proccss.
9.3
Intcrcalibration
It was decided by the planning group to utilise as many oppartunities as possible far interealibration during the
1997 surveys. In order to minimise the effeet of spatial and temporal variability of hcrring abundanee, the
exereises are intcnded to be inter-ship calibrations, with the vessels running the same course at the same time.
Since sueh an arrangement will require some extra time for eruising, whieh will inevitably reduce the coverage of
the sampling area to some extent. This was judged to be aeeeptable.
•
The anticipated area for the first intcrcalibration is around 58°N and ODE. Thc vessels sehcduled to meet in the
area are GO Sars, Tridens and Walther Ilerwig Ill. Thc Walther Ilerwig wiII leavc port 23 Junc and wiII sail to
its sitc far eeho sounder calibration, presumably Kristiansand, Norway. After gear calibration she will saH to
58°N 02°E and will start surveying thc area by covering transeets in E-W direction up to 02°W. From then on
shc will eover transects 15 Nm apart in northern direetion. Until thc anticipated day of Intercalibration (l July)
Walther Ilendg will have eovered about eight statistieal rectangles with probably relatively high fish abundance.
Thc exact location of meeting for interealibration will bc determined after thc area has been serutinised and will
bc communieated to the other ships from Walther Ilendg by radio. Radio contaet will bc established prior to the
meeting at 10:30 UTC.
Thc vessels Tridens and GO Sars will attempt to reach the meeting point in the morning of 1 July. Thc
Intcrcalibration will be carried out throughout thc cntirc survcy day, during whieh no fishing will take plaee.
A sceond ealibration will be attcmpted bctween Walther Ilem'ig and Dana after the complction of the first
interealibration. During 02 July Walther llem'ig wilI sail eastward. Radio contact between Walther Ilem'ig and
Dana will bc established 2 July UTC 1030 for agreement on the precisc loeation of the meeting. This will be in
the early morning of 3 July on thc antieipated position 57°30'N and 06°00'E.
Further intercalibrations are antieipated between GO Sars and Scotia on/about 16 July and between Scotia and
thc wcst coast charter on/about 26 July. Details on timing and loeation will be arranged by radio contact betwcen
thc two ships.
Details of various ships communications are provided in Appendix G.
•
9.3.1
Proccdurc for the intcrcalibration of echosounders during thc North Sea herring survey
Thc vessc1s should bc positioned with onc in front and the other 0.5 Nm behind at 5 on thc starboard sidc. When
three vessels take part simultancously, the third vessel position will be 0.5 Nm behind thc leading vessel, at 5 to
the port side. In this situation thc second and third vessel are stcaming parallel.
Thc speed during thc Intercalibration should be 10 knots or adapted to the vesscl ....ith the 100vest praetical
integration speed. Thc integration should last for at least 12 hours. Due to the very limited time period, the
intercalibration with Dana is restricted to 40 Nm.
The vessels takc their relative positions and start sailing at thc agreed speed and course. When the vessels are in
a stable formation the, the leading vessel gives a start signal and starts his own logging. The other vesscls start
their logging after steaming 0.5 Nm. A synchronising signal should be given by the leading vessel every 5 Nm at
whieh time all vessels should record their geographie position and annotate their eehograms accordingly. Thc
lcading vesse! should be ehanged frequently ensuring that eaeh eonfiguration is earried out at least twice during
the proeedure.
.
A sampling interval of 1 Nm should bc used for integration. Thc integration should start at 10 m bclow watcr
surface and the SA-values should preferably be stored by 10, 15, 20 or 25 metre layers depending on thc
E:\PFaPGlIERS97\REP.DOC
01/04/97
7
intercalibration area so that 10 surface ·channe1s can be registered on one echogram. Threshold for the echogram
should be set to -70 dB. Normal survey settings should be used for all other parameters.
9.4
Exchange of length and age data from trawl hauls
A study of the length frequency and age proportions of trawl hau1s by different vesse1s in similar areas is to be
undertaken. This study will take place in Aberdeen and results will be submitted to the HSPG meeting in January
1998. The data will be collated from all hauls which containcd hcrring undcrtaken during the North Sca and Via
surveys of 1995. Each partieipating country will submit their data to Abcrdcen by thc cnd of April 1997. The
format for exchange of data was discusscd and is based on the exchange specifications for the IßTS data (Anon,
1996b). The agreed platform for data exchange was a spreadshcct in f\1icrosoft Excel '15.0; a blank template
spreadsheet was supplied to all participants (xxx_95tr.xls - participants shou1d savc data filc as "leES country
code_95tr.xls"). A table detailing the entries of this spreadsheet is appendcd (Appcndix F). The proccdure for
the analysis of the data will be determined in Aberdeen.
.
10
RECOMMENDATIONS
The planning group recommends that:
1. Duc to inconclusive findings in an examination of the herring survey time scrics that furthcr studies be carried
out on:
•
•
•
•
~
the separation of west coast and North Sea herring stocks within the acoustic survey time series;
the dcpth related distribution of herring and its impact on the stock cstimation;
the use of GAMs on acoustic and IßTS survcy data;
an examination of prc-1991 surveys for possible under estimation due to signal saturation in the
electronics.
2. The acoustic surveys should be continued with each participant covering the same general areas to maintain
consistency and a number of stcps be taken to improve qua1ity assessmcnt in the acoustic surveys:
.•
•
•
surveys will inc1ude inter-ship calibration;
a study of variability of trawl performance bctween participants be carried out;
a workshop be held in Bergen in January 1998 to study variabi1ity in echogram scrutinising procedures
between partieipants.
3. For the larvae surveys:
a) Yearly surveys shou1d focus on the southern North Sea as weil as on the Orkney-Shetland and/or Buchan
area. A more dctailed analysis of the historieal database is rcquircd to elucidate whieh of the two northcrn
arcas should rcccive a higher priority.
b) Efforts should be made to organise for a complete coverage every three years, out of phase with the
mackere1 egg survey, starting in 1999.
c) The effcct of survcy timing on 1arvae abundance indices and production cstimates should bc examined in
more detail from thc historiea1 databasc, to confirm or disprove thc indications so far available.
d) Reliability and changcs of thc zIk valucs should be studied as the LPE is espccially sensitive to this
parameter. A standard procedure to estimate zIk should be dcfincd and thc existing data serics reviscd
accordingly.
4. Thc planning group for herring survcys should meet in Bergen, Norway from 19 to 23 January 1998 undcr thc
chairmanship of E J Simmonds to:
a) Coordinatc the timing and area allocation of and methodologies for acoustic and larvae survcys for hcrring
in thc North Sea Divisions Via and lIla and the Western Baltic;
E:\PFaPGHERS97\REP.DOC
01/04/97
8
•
b) Combine the survey data to provide estimates of abundance for the populations within the area;
c) Hold a workshop on acoustic echogram scrutiny;
d) Assess the results of studies on: the separation of west coast and North Sea herring stocks within the
acoustic survey time series, the examination of pre-199l surveys for possible under estimation due to
signal saturation in the electronics, the inter-ship calibrations, study of variability of trawl performance
between participants.
e) Review the results of the above studies and then report on the applicability of a furt her study of the herring
survey time series.
11
REFERENCES
Anon, 1996a. Report of the Herring Assessment Working Group for the Area South of 62°N. ICES CM
19961Assess: 10.
•
Anon, 1996b. Manual for the International Bottom Trawl Surveys. Revision V. Addendum to ICES CM
1996/H:1.
Anon, 1990. Report ofthe Working Group on Herring Larvae Surveys South of62°N. ICES CM 1990/H:32.
•
E:\PFaPGHERS97\REP.DOC
01104/97
9
North Sea Herring
2500 .
Acoustic Surveys
'f
I
2000 ...
,,
I
,,
I
~
I
.....,
Cf)
.~. . w
Cf)
Cf)
I
1500
I
ca
I
.
E
o
.OJ
_
o
O'l
C
,/
,
..
1000 --
/
.
'c
"
.
. .
...
.
WG
Forecasts
:',/
,
...
"
/
3
ca
c..
Cf)
500 .-
~
...
...
"
--.
.....
1995 WG
Assessment
o. -·-----t---·---····--j----r--·--+-------f---t--------Jro
...
~
00
w
00
~
m
o
m
00
00
00
00
m
cn
"r""
"r""
"r""
0')
cn
cn
m
"r""
~
~
~
m
--/N
C"')
I ·----1
oq-
LO
m
m
m
cn
cn
cn
cn
"r""
~
~
~
Q")
Year
Figure 1
Discrepancy between the predicted and estimated spawning stock biomass. Solid line represents the spawning stock biomasses
estimated at the 1995 Working Group meeting. The broken line "WG Forecasts" represent the forecasts made by the Herring
Assessment Working Group from 1982-1993. The "Acoustic Surveys" have indicated a spawning stock biomasses over the period
1989-1995, wh ich are double those calculated than those calculated by the assessment.
•
..
12
APPENDICES
A) Amplitude distributions for Scotia surveys in IVa 1987-1996.
B) Abundance, and Biomass of herring by ICES statistical rectangle from acoustic surveys 1984 to 1996.
C) Comparison of acoustic and IBTS times series by examination of relative cohort strength.
D) Perceptions of North Sea herring stock dynamics and survey variability that are robust to catch misreporting.
E) Effects of reduced sampling effort on abundance and production estimates from North Sea Herring Larvae
Surveys.
F) Format for exchange of trawl sampie herring age and length data.
G) Communication information for research vessels.
H) Planning Group for Herring Survey contact numbers.
•
E:\PFc\PGHERS97\REP.DOC
01104/97
11
APPENDIXA
AMPLITUDE DISTRIBUTIONS FOR SCOTIA. SURVEYS IN IVA 1987-96
Background
The following analysis was designed to examine the amplitude distributions by both quarter ICES rectangle and
Elementary Distance Sampling Unit (EDSU) for the time series of Scotia surveys in ICES area IVa. The aim was
to determine if there had been any dramatie changes in the performance of the surveys which might explain the
alleged discrepaney between the acoustie index and other indices used in the assessment
Methods
Core area
A core area for the Scotia surveys was designated, based on the 1991-1996 surveys. A quarter rectangle was
inc1uded in the analysis only if it had been surveyed in all these years. Any rectangle missed in one or more
years was dcleted. The core area is illustrated in Figure 1. For surveys prior to 1991 rectangles were included
only if they were within the core area. However for some of these surveys coverage of the core area was not
complete. The EDSU data set was filtered in the same way to inc1ude only EDSU from within the eore area to
allow direct comparison between years. The result of this approach is that the biomass and abundance data
presented are not exact matches for the figures reported for the survey in a particular year.
Data sets
The biomass and abundanee for each rectangle and year were extracted from the ICES coordinated survey
reports for the area or from individual survey reports prior to 1989. The EDSU data set are echo integrals per 15
minute sampling period and were extracted from digital data recorded during the surveys on Scotia and used in
the subsequent analysis presented to the HAWO. Following extraction the integrals were corrected using the
echosounder calibration data for each particular year.
AnalJsis
Biomass and abundance by rectangle
The data were sorted into bins (classes) for presentation as histograms. For abundance the bin size chosen was
10 million fish and inc1uded a zero bin and greater than 200 million fish category. For biomass the bin size was
5,000 tonnes and inc1uded a zero bin and greater than 100,000 tonne category. The data are presented as
absolute numbers of rectangles in each c1ass. It should be noted that the earlier surveys inc1uded fewer
rectangles.
Echo integral by EDSU
The data were sorted into bins (c1asses) für presentation as histograms. The bin size chosen was 500 and
inc1uded a zero bin and greater than 10,000 category.
Results
Biomass and abundance by rectangle
The histograms for the 10 years are presented for tonnes by rectangle (Fig. 2) and numbers by rectangle (Fig. 3).
A number of changes can be seen over the 10 years. For numbers (Fig. 3) up to 1990 the frequeney distributions
were relatively flat with similar numbers of rectangles in most of the lower value bins, less than 10% of the
rectangles surveyed contained no fish. Distributions were more skewed for the biomass data in these years. One
possible conc1usion is that a lot of the fish in thc middlc range abundance bins were relatively young and
contributed )ess to the biomass va)ues. Following) 990 the distributions were much more skewed, with the
number of zero rectangles generally being between 20 and 30% of the total. It is interesting that in all years there
E:\PFaPGHERS97\REP.DOC
19/03/97
12
•
were small numbers of rectangles with abundances greater the 200 million fish and that this did not seem to
fluctuate much with any change in stock levels. To illustrate some of these trends the frequeneies in the zero and
lowest value biomass bins were plotted against year in Figure 4. The increasing numbers of zero rectangles can
be seen, and a possible decrease in the number of rectangles with low biomuss.
Echo integral by EDSU
•
The histograms for corrected integrator values are presented in Figure 5. Again as in the rectangle based data
there is strong evidence of an increasing number of zero observation over the time period - between 30 and 45 up
to 1990 and generally greater than 60% thereafter. Figure 6 illustrates these trends in more detail with the .
frcqucncies in the zero and lowcst value bins plotted against year. Although there are some fluctuations there is
dear evidence of an increase in zero sampIes over the period. However this is strongly mirrored by a dccrcase in
the number of sampIes in the next bin. The most likcly explanation for this is that over the years, the operator has
attached less importance to very small fish schools on the echogram. This possibility is supported by the general
perception that the most important element of the biomass is contained in the relatively few larger sampIes, and
that it is not worth expending cffort allocating vcry small traces to species. It should be noted that this would be
expected to lead to a small undercstimate, as some herring schools will be missed, however, it is unlikely to bias
the stock estimate upwards. Figure 7 show the percentage of zero sampies in each year plotted against the
biomass in the core area in that year. Interestingly there is a possible trend of increasing percentage of zero
sampIes with increasing biomass. However, this is likely to be seriously confounded with the observed change in
operator practice noted above.
The same data are presented in a 3D plot in Figure 8. Apart from he increase in zero sampIe frequency, there is
!ittle obvious difference. It may be possible that there are more high value observations later in the period. To
c1arify this the percentage of high values against year is plotted in Figure 9. There is no c1ear overall trend.
However, the level is fairly stable to 1990, a sharp dip between 1990 and 91, and following this a possible
increasing numbers of high values from 1991 to 96. Figure 10 shows the percentage of high values plotted
against biomass, and no dear relationship can be seen.
The final plot (Fig. 11) shows cumulative frequency distributions by year. The data have been normalised and
only the last 200 points plotted representing the maximum number of non zero sampIes in the core area in any
year. The only obvious pattern in this plot is that again, the years 1987-1990 are c1early separated from the later
years, having generally shallower trajectories.
Discussion
•
The main conc1usion from the study is that, at least for this survey area, no obvious explanation for the alleged
change in survey performance over the last 10 years can be seen. There is some evidence of a change in the
performance ofthe surveys between 1987-1990 and 1991-1996. The most likely explanation for this lies in the
changcover from Sirnrad EK400 to EKSOO echosounders. The dynamic range of the EKSOO is significantly
.greater than the EK400, and it is possible that particularly dense schools resulted in saturation of the system.
This would tend to reduce the amplitude of the high vales and may have resulted in a tendency to underestimate
in these years. The perceived tendency for more zero sampIes over the period can largely be put down to
operator changes, and would be expected to result in a slight underestimate of the stock.
E:\PFaPGlIERS97\REP.DOC
19/03/97
13
4W
o
2W
2E
62N
61N
BON
-+.
>
~
-" "
.,"'-'"(
!
'j
/
.
.i
>.
.'
~."
59N
-
58N
~
i
Figure 1
f-----t:.'/~---f----+_~__l Core area used in analyisis.I--f-_ _
--1
"
~-..._~
100m
..,....,.\
Scotia 1987-1996
V
....._--"'"---.Aberdeen
J
•
t,
igure 2. Numbers by rec_ngle
Scotia 1987-96
50 -r----~---,..---~. -----,---,-,-,
40
r
O
-
t20
o _~.~
: •
50
40
r
O
t20
50
-iJI11993
40
1
10
o
11 ~r .. 11 '21
.. ' ..
..
'11)
fI'" ' .
"'''tI
,...".,.,.
1111'1"
I.·'"
Ifll'lf.
, .... t!lO_'
tf.O.~
:.-__.J__'__ .'
10
20;
')0
.. :..
7f
ID
•
'100 HO
to..mb...
:
111 UII' 140
+--1--
t~ :,lfl
I
111' "" ,,,:lOII
Figure 3. Numbers by rectangle
Scotia 1987-96
30
T---r--'T1987 I
20
'
,
1: l!litiJWli,~
• :,f.oo,,'iC1 Jll~"'~':eo
30
30 .
7O!1lI:. 100;1l1111l~'" 140:"
20
i
01:
10
o ,.
.:'71 1.:'110. 1
1·1]'' JL:
. . i .: ! . . ": !; [ . i
~
:.,f..;.
20:"
4lI:'J1
.!,.
111
~.
~i-,
(11993
30
1':'.
lfO:'\Ill11l11:lOD
!
i,
11'·1•
UJ
0"
Nl.mb".
110 11.:140
TI-I -- i-111199i
1
( 1-
, _ ~ . c
Ni,oe 1ft
'IE.QlJ''IO
. ' • • :",
.1.. .~ . .!.'
6O'n
••
.
I
20
i
:
_
"IllIUOUOf)O:'''I'JI:,.,,,:,.t.t 200
.~
.r20
I TT-iT--i-----1~
"10
Nl.mbe<.
30
30
0-,.
,
~f.oe 'NI:. ''':
:
40
i
,.!. ": 10
.
10
'.-H' UO '"
Nl.mbe<.
i!
",!"
:~t7ll
.:
'.;110
ao!
'~'-I '1996 "
i
ot
•
10
0
I
11 .. ,.
11
•
..
..
•
'"
..
..
,.
NlA'tllJers
110
111 ,..
I. t1t
,.
'"
110 ,_,
Co
.'.
'
e
·.
e
Figure 4. Percentage of rectangles in bins 1 & 2
Numbers - Scotia 1987-96
1988
1990
1992
Year
1994
1996
Figure 5. Corrected Integrator frequencies
Scotia 1987-96
60
::
50
"~ 30 -~
J: 20
10
o
50
1987
1
~40
er
60
::
I
1
'-+-H-t
~r-j:::.=Li--[-i
1-' t-.. ._.) '-'---TT"
-1-'--r'r'
"'1--'("-"- ;
~i~-t--
-------_..
. :.. ;.-_.;. . r----····-r-·---
~40
"gJ 30 '.
0"
3: 20
10.
o
• :lI«alO...., .... ~ ....".__..._ ...., . . - _....,.._,.... _~I-.1
Corrected Integrator value
60 - .-~~-,--­
60,- __-.,......,-
50 -
50 -.
~40
~40
..
g"J 30_
.
gJ" 30 -
0"
0"
J: 20
3: 20
10
10
o
o
' . ~.SlGDO.,..,·. . . . . ~·. . ~ . . . . . .•. . . . . . . . . . . . .· _ . , . .. . . . ._
. . llDIIaIl
Corrected Integrator value
00
60
~40
1~94
50
1993
1
":-l-l------rr-- r
60
: -----:-.:
50
1
"~ 30
g40
J: 20
3: 20
i
~30
er
~
1 _
i
i
f
j
::
c. I,_·~·_t_-_-_-_-'_~_;__:-_-_- _-r~·. - ;-\_.-_-_._~_ -.·-·---.,f-----'----·
•
f;:-I~C--!rr----r!
!:
!: :i
OLJOLJ·.'
60
;..!
,_-_-.,1_·
..
O~.~. . . .JIlIII~..... :mI. . . . ._·~·,.__ . . ·~·~_. . _lIWalOllllOOo
Corrected Integrator value
! :: _
i.mMl.~.tS.
~40 -;
"gJ 30
fJ
J: 20
10
•
Corrected Integrator value
. .~
•
,.
_
,..
Corrected Integrator value
60
••.,..IG..-.......a.._.,;............,.._. ..,...~......_
1
..
..
.z_i.
50
o
1995
.· Lt_-~1_ ~· _ I_- '_- _-; I_- _- ._~;. !-_ -_ ~I.,-: ,~I-. ;~_-'_-;i-'·"TI~
OL. .
:
•
1
..
_
_nal~ •
e
'.
e
r------------------------:::;~--__._--.-------___,
Figure 6. Percentages in first two bins
Scotia 1987-96
100
-r--..,.---~--r---.,.---_.___...--,-------r---r----r-----,
4---t---t----t-------~W""---Ttl-
I
I
80
-1----+----,
I
c:
--c--
-.---- ... -----.- .... --j-- .. ---.--c----
•-
.S
60 _-----------1--Q)
I!
••••
C'O
c~
.
~ t··~·· _
40 -. ------ -
Ci>
.~~
i
--.. -~ ,,- - -.- _.__
1
.I
L
I
II
L
~~~.~~....• ;
/
I
•....
!
i
I
·..·····-_·-_···-··11_ -.-.-- ..-
I
86
. . . ~•
.... _.
._.
_
•••••
_..-
}'"
-
:~ •....-.····4l.
--
~:~ ~.:-_
r-=-:-~---,
I First bin
,
:;;.4I~
.
I
I
I
88
......
- ···············-······ ···f--..--··_- ------.----..-.-..- ---- - -"""-"'-'-'-'
o- I · !
,
,A~
•• ' I'.
20 ---- - · · · · · · · · · · · ·. 1······_··..-·__····· - _ - ··············..····..···-t····-····-····_··.L-·····..····- _.. . .
".
-------
I
!
I
-.J._ --.---~ - --..-.~-- -.- --.- -.- -..---- -.--..-._--..
••••
1(
+_..__..__.. . --j._..._ .
········ ..··· ..
-_
".
1/
~
I
·····1--..·· · ·. ··_·_·_- _ _ · · · · · . · · · ·-·.. · · ·. ··········-····t·····..··~~·~.::~: . • ::~
I
I
I
f--
~.
1
I--·----·-t-·-·-----·j-·---Jf---·-·--
..
".
/
i
l
!
.
.
.
.
.
._ .._.._._::::-;.::~ __.
'__'~:":_"" __...!'__.. j__._.
..
--._1 Bins 1&~
rO.-------~=-=.=~L
. . --L-:-.-L----.1
Zero bin I
!
...
I
1
a.
I
~
'11'"
i !
!
.c
C')
!
~
90
92
Year
94
96
I
...
Figure7. Percentage of zero sampies against
biomass in the core area 1987-96
80 .- . - - - , - - - - - , - - - ; - - - - , . - - - r - - - - - , - - - - - - - ; - - - - - , - - - - - , - - - - - - ,
~ 60
0
I
~
0
i
i!
I
·_·_·~
_\
If
SI
~
- ._
_
l1li
i!-T----------r---l
•
,
!
t
i
I
i
i:
-
...........................--
!..
I
!
I
"s
i
i1
~
-.-.---;----
!
:
i
ii
I~
i
J
_
J_
1
I
!
Ii
i
!
I
I
[
I
,
I
,
:
j
I
i
•
!
40
_._ - _ -i _
30
.....1....
- __
+I . -.- _
!
- - - - - - - - - - - - -..r ..· · - - - - - ·
1
i!
;
j-
II
i
!...
!......
i,
l
I
I
I
t
, ,
I
f
.:.'
i
I
;
:
I
.L._
--.--
1._ _.._..__ _ _l--.__
!i
:I
!I
-
I'
J
t !
_
__ ..-t -
,---
1
-
!i
II
:
.. ,
i
I1
1
_
i
iI
--
I
"1'
:
....................................."'
~50
0
I"
__·_H.H..·t·..- - - - - ·· ··..·- t"..-- ..
;
(1)
0
iI
..---.--.-.l -.-."'..-.-.~- ----~--.- - _.__.1..._ _
i
i-i
!
............... !
i
~
»
·_·__··_·..·t··
L.
N
N
I
- ·
I
11
Ti
i
70
I
,
"' l...
I
!
!
i
- _
! :
"1"
i
i
_
l - - - --l-.- ---
:
'
I
i
;
·
·r..·.···..·
--
I
I
_ _ )...
,
I
;
I
!
j··
I
~-.-.
I
_...
i:
1
:
:!
i
i
-
'I!
:
!
I,
1......-
I
I'
:
:1
i
i
: !
i
! '
- - - !- _ .
I!
I
I
i
I
_
I
-----. -
!
I
-ri
· · _ ··
..._
- .
i
--.........i..- - -._.
i
---+------t---r--··---!-_·~-_··-+--t--t----- ----1-
400
600
•
' :
,
800
1000
Biomass in core area
•
1200
1400
..
,
Figure 8. Percentage by class in core area 87-96
80
60
40
IV
20
o
'------_._---------_._---_._-------------------
.. -----_.
---- ---_._--------_._-----_._---_._--_.-
- - - - .. -
_. .
.
Figure 9. Percentage of high values against year
in the core area 1987-96
14
. . . . --c---------~------_,.-------_,__-------_:__------,-__,
I
I
1
........... ;
)
l
I,
''
,
i,
·········f··········
i
I
:
~
j
···············..···r···························..··· .. ·...
:
I
:
I
+.........
-
··+·..·-····..·-· - · - -1·-_····--..·-·-······· -._ __._ _+._____..
'
I
I
I
I
I
i
i
I
"
.1-_
;'
:
I
12 . . ---.. . . -~ ---..-- . . . .-. . "l" . .----.. --- ----r- -··---------1---- -----·----t-------------,---------- --T----- ----1----------1---------- 1---;
..
tn
Q)
-co:J
10
N
N
I
N
1
1
J
i
....-.-
L
_._ l
_
-..-.,
I
!
-'=
C)
. J
i
,
. . . . . . . . --;---
>
»
.....
1
_
--
+._
'
_.+i
.
.
',I.
_._
1
I
.1. -··-·····-1···---· ---····1-·-·- -1·1..-
-··.·····
i
- -..j _I
I
1
·.. \-··- · ·.········ ·.·1·- · - -
I
.j..- -.-.I
+.- _--- - -_ -- ! --..-- -.+.- _-.
I
I
!
I
-j·. ·- ·..-·- · -. - -· ·f--·-----·---- --.- -.~.-.- ---..---i----.
I
I
i !
i
i
1r·-··
_ -.-.,..:
.
- :i ·····lI ····· · _··i·1 ·_· ··..· ·..····· ..· I
8
.I
i
I
-'=
. ·.-. -..-+·-·.. --..· . ··--·-1·-..· - - - ·.. . ·..··---+.. -·· ·--·..·- . .·--1..·.·.---.. .-.-.. ·..··-·[ . ·.. -·----· .-.-+.. --..'+0
;
I
!
I
i!
6
.
·
···.
.
....·r·
/...
-.:
j
-.
..
.
)
1 - --. I
~
0
t-.j
,I
8
11--_ __II1II'
;
. . . . . . . . . . 8 . __ ••
1!.. _ .~ ..__.
I
.
4
2
,
--+-_..
I
····1-I
I
I
I
t·····---··-------.. . - ·.._-_······t··_-_..·······..__..__.._-_··.. ,· · . ··_-_.:.·
:
I
I
I H... i . . . . ·····1····..·_ . __ . ·
' i ! '
i
I
i
I
i
····!·..·-.····---..---------·····--··..
I
I
-t .._- ~
'
8'9
90
···-·..· ·.._· ·..tI· · ·..··
i
i
. -.r--.-.. .----+-..
I
--.J- - -.-.. --y
J
- -
I
I
----··_------_···-------···'r·------------·-----------..--]..------..
------------------j-----------_.------....
'
I
i
····--1
I
!
'
1
.._.. . . . _..L.................
i
!
_.+
:
i
--
I
---l----ui
i
!..· ·- . ·.. ··-··I·.. .---..··,·-····..Ii··,----
._.f-- _._~--+--
I
; '
~
i '
I
i
!
~:
-1---7---l----+----t---t-----+----+---f----l----!----J
8'7
88
I,
1
91
92
Year
1
93
94
i
9S
96
.
' ,
•
Figure 10. Percentage of high values against
biomass in the core area 1987-96
600
800
1000
Biomass in core area
1200
1400
I.
Figure 11. Cumulative Frequency plots by year
- last 200 points 100
80
~
(,)
c
Q)
~
C"
60
Q)
''tQ)
>.....
I
N
+>.
.)::..
>
+:i
-co
~
40
E
~
u
20
87 ;,~;,,,~:'...:':'A'.'.i",...:,.<jMiii!U
92
r;.::.::.
- .
{:
\."".;,.
88 - - . 89 - - . 90
93
;,;,;-t.,'
.,
'"
,;,;.:.:.:,
94
""':.;i-':':":;.\:~
•
';V1.':
95
..•••....
~''.~
,'/"',,':"
..
·::i.'
91
96
----
----------------
APPENDIXB
COMPILATION OF THE COORDINATED ACOUSTIC SURVEY
TIME SERIES ESTIl\'IATES FOR HERRING
Marked variability in the annual estimates of abundance and biomass of herring stocks in northern European
waters has contributed to the requirement to review survey design and data analysis. One component of this
review has been to construct a spreadsheet application which will provide estimates of abundance and biomass
from the existing time series by age and ICES statistical square for both the North Sea autumn spawning stock
and Baltic Sea spring spawning stock respectively. This was undertaken in Dctober 1996 and has recently been
completed.
Using the annual ICES reports for herring acoustic surveys, the proportions of abundance at age within each subgroup in each cruise were used to ca1culate abundance and biomass (incorporating mean weight at age) on the 30
mile ICES statistical square scale. The resulting abundance and biomass estimates were then checked against the
estimates published in the annual survey reports and combined. For illustrative purposes the abundance
(millions) and biomass ('000 tonnes) estimates for mature autumn spawning herring by year are given in Figures
1-13. The annual abundance data for mature autumn spawners are also presented as contour plots for each year
from 1984 to 1996 (Figs 14-26).
Compiling the individual reports from several participating nations was not straightforward. This was due to
differences in format and presentation. In order that future surveys can be combined efficiently it would be
helpful if participants could provide data in a standard format. Specifically, reports should include:
•
A map of the cruise track overlaid with a grid corresponding to the ICES statistical rectangle scale. This
should be appended with the number of 15 minute integrator runs in each rectangle.
•
A corresponding map showing abundance (millions) and biomass ('000 tonnes) together with the boundaries
between the sub-areas (strata).
•
Two tables giving herring abundance, mean weight and biomass by age (0 to 9+), maturity and sub-area for
North Sea autumn spawners and Baltic spring spawners respectively.
•
E:\PFaPGHERS97\REP.DOC
19/03/97
25
Numbers (millions) of mature autumn spawners (1984).
63
-,..
62
:So'
~
"-
61
60
-+-----+---+-----+----+---+-:,...:::,.;;;~
59
-+-----+---+-----+----+---+hLH--tlL~2'8;w::::~_+---+_--_+_--~
58
-+-----+---+-----+----..;~
57
-+----t----+-----t-~---'....-.n
w
o
:::J
I-
t=
::i
56 - t - - - - + - - - - - t - - - + - - - - L J I >
55 -+-----+----+----,;
54 - t - - - - + - - - _ . . i i
53
-+----+-----t~-
-14
-12
-10
-8
-6
-4
-2
o
2
4
6
8
10
12
14
LONGITUDE
..
e
,.
Numbers'millions) of mature autumn spawners (1985).
63
s:
-~
62
~
"-
61
60 +----+-----+---+----+----+----'<:----=
59
+----+-----+---+-----+=-----.:..:..:.+--='l~-++
58
-+-------+------+----I---~
57
+----+-----+---+-...-~r-f"
w
Cl
:J
I-
~
:)
56 - 4 - - - - + - - - - - t - - - + - - - ' - 1 I ,
55
-1-----+------1---,;
54
-1-----+-----,,;:,
53
-+----+-----+~
-14
-12
-10
...
-8
-6
-4
-2
o
LONGITUDE
2
4
6
8
10
12
14
Numbers (millions) of mature autumn spawners (1986).
63 +-----+----1----1----1------+----+-----+---+----+---+--::
62
- l - - - - I - - - - - + - - - j - . = : . ...--+---+------I----I-----+---+--_.._=;
"!t
\.
61 -+-----1-----+---+------1----1----:---+-;----+-----+----1---41:.
60
+---+---j---t---r---j-"iiit~
59
w
IV
00
Cl
:::l
t::: 58
I-
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
100
50
0
<t
-'
57
56
55
54
53
·14
-12
-10
·8
·6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
Numbe' (millions) of mature autum' spawner~ (19~7).
63
-,.. :s:
62
~
\.
61
60 -+----+----+---+-----t------t-
59
w
0
IV
'0
:::>
I-
i=
58
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
" 100
"50
0
:5
57
56
55
54
53
-14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
Numbers (millions) of mature autumn spawners (1988).
63
-'lltoO:
62
:s:
~
\.
61
60
--+----+----+------+-----+----II----*+I+-I-~~-*_--_+_-..."...,......+_-~
...
59
w
0
w
0
::>
I-
~
58
.,
1000
950
900
~
57
850
800
.
750
700
650
56
000
550
500
: 450
400
350
55
54
53 -+----+----+--'::..-10
-14
-12
-8
-6
-4
-2
o
LONGITUDE
2
4
•
6
8
10
12
14
Numbeft (millions) of mature
autum~spawners'(19~9).
63
-,..;
62
S ..
~
\,
61
60
0
59
w
w
0
:J
t- 58
~
"
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
300
5
.
57
56
55 +----+----/---.'-
\@i~~~
54 -t-------+---...,
,""" 150
",;;,;: 100
50
o
53 +-_ _-+-_ _-+--=:J_
-14
-12
-10
-8
-6
-4
-2
o
LONGITUDE
2
4
6
8
10
12
14
-
-
--
-
--------------------------------------
Numbers (millions) of mature autumn spawners (1990).
63
-'lQO
62
~ ..
,.
~
61
60
-t----t----t-----+----+---f---,4";,,I,~~_#;~
59
w
C
w
N
::J
I-
i=
58
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
',$'".300
'<
250
:3
.
57
56
55
-!----f----I-----,';
54 - t - - - - t - - -__
ii;:I
00'00
53
-12
-10
100
50
o
-t----t----+~­
-14
~~~
-8
-6
-4
-2
o
LONGITUDE
2
4
6
8
10
12
14
I
t
I
Numbers (Iillions) of mature auturnn sp'wners (1991).
63 - 1 - - - - - + - - - + - - - - j - - - - - + - - - - + - - - - + - - - - - f - - - - + - - - - I - - - f - - - - - - ; ;
61 - 1 - - - - - + - - - f - - - j - - - - + - - - - l - - - H
60
10e
59
90e
w
c
w
w
:J
!:: 58
80e
<
..J
loe
r-
57
60e
50e
56
40e
55
30e
20e
54
10e
53
-14
·12
-10
-8
·6
-4
·2
0
LONGITUDE
2
4
6
8
10
12
14
0
I
Numbers (millions) of mature autumn spawners (1992).
63 -1-------1----+---+---1----+----+---+---1-----1----+---,
62 -+---+----+---+-~
61
h--+----+-----j----+---+--_+_~
+---+---+------1,-----+---+--;r-4-~~~~--+-_____\,.--+__tI;.g
60
1000
950
900
850
800
750
700
650
600
550
500
450
400
.,,,,,,
350
~~
'.~ 300
250
200
150
100
50
0
59
w
c
...
w
::>
!:: 58
t<t
-J
57
56
0
55 -
.... ;.
;,:;,
54
53
·14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
e
.
'
Numbers'millions) of mature autumn spawners (1993).
63 + - - - - t - - - - - - I - - - l - - - - + - - - - - - + - - - - t - - - + - - - + - - - - - I - - - - + - -___
61
-+---+------l----+----+----+---==-''4-----.I--+==+-t~+------I___tJ;:
60
1000
950
900
850
800
750
700
650
600
550
500
450
400
. 350
300
250
200
150
100
50
0
59
UJ
0
w
Vi
:)
!:: 58
....
«
...J
57
56
55
54
53
-14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
-----------------------------------------------,
Numbers (millions) of mature autumn spawners (1994).
63 - + - - - - + - - - + - - - - - f - - - - + - - - - + - - - - - I - - - - - - - - j l - - - - - - - j - - - - f - - - - t - - - - - - :
62
-+----l----I------t-----"'" h.---l-----f----+----+-----+-----I----V'
61
-+---+----+----+---+----t-----;4-I-7-''------i1H------I----t---4~
60
59
w
c
w
0-
.-::J~
58
ct
..J
57
56
;~-
55
~
~~
<
...
54
53
-14
-12
-10
·8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
\
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
100
50
0
e
Numbers
.
(m~ons) of mature autumn spawners (1995).
63 - l - - - - l - - - + - - - + - - - + - - - + - - - - j - - - - - - f - - - - f - - - - - - l - - - - \ - - - ,
62 - l - - - - l - - - + - - - + - - " ' "
61
t'r--l-----j-----j----+----+----+--..i"
+-----l---+---+---+---+----=-t---i-H+~H-+--t--------l----'''iiIII
60
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
300
250
200
150
100
50
0
59
w
w
.......
c
::)
t: 58
tel
-J
57
56
55
54
53
-14
-12
-10
·8
-6
-4
·2
0
LONGITUDE
2
4
6
8
10
12
14
Numbers (millions) of mature auturnn spawners (1996).
63 - I - - - + - - - / - - - - - - ! - - - - + - - - f - - - - + - - - - I - - - + - - - + - - - f - - - - : ;
62
-+---+----l----l-~
R;-+---/------!----+---f----+-----v,
~
61 - + - - - + - - - I - - - - l - - - - + - - . . . . - + - l M +
60
o
-+---+---I----l----+-..,--~
1000
950
900
850
800
750
700
650
600
550
500
450
400
350
300
250
:,200
150
100
50
w
o
w
00
:J
t: 58
-+---f---+-----t'-e-~~
~
..J
55 -t----+----f-----,:,
54
-j------If----.i
53 -+---l------+---==-·14
·12
·10
o
·8
·6
-4
·2
0
LONGITUDE
2
4
6
8
10
12
14
e Mature autumn spa,vners ~984).
63
"
ABUNOANC
~~'
-'lI!lO'"
BIOMASS
62
~
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
23.10
5.76
0.00
0.00
1.27
0.25
0.00
0.00
0.00
0.00
0.00
0.00
63.42 11~~" '-l~4.70 10.00
14.68 2
,. 53.05 1.95
0.00
0.00
5.80
1.45
0.00
0.00
0.30
0.07
16.60
4.14
0.00
0.00
\.
61
5.20
1.25
60
0.00
0.00
>0
i=
178.46 54.73,t2 '1l7 138.89 21.37
35.~8 10.~. 76 22.68 3.47
3.51
0.58
0.00
0.00
0.00
0.00
0.00
0.00
o.ott:. 4.06 42.09
0"6.94 6.45
23.92
3.92
0.00
0.00
10.65
1.56
0.00
0.00
20.43 244.39 28.53 54.96 13.05
2.95 36.18 4.21
9.02 2.14
0.00
0.00
0.00
0.00
48.85 372.11 0.00
7.10 61.04 0.00
0.00
0.00
j ..~~ ••• /;f·
1J~:X
../fA
1
W
C
.....
.
64.15 134.21 519. 1 147.72 0.26
11.14 19.18 86.69 33.17 0.05
l
59
:::J
to-
32.00~.06
7~ " .14
58
..
'
:5
8.16
1.30
0.00
0.00
14.83
2.23
57
0.00
0.00
"
56
55
.,'
54
53
-14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
-
-
--
---
----------------------------------------------------------------
Mature autumn spa,vners (1985).
63
ABUNDANC
~~I
BIOMASS
-'liIIro!
62
':t
'-
61
0.00
0.00
5.6t
1.19
60
34.34 6t.91 ~7.51 223.00 0.00
7.29 13.Jl11.57 44.08 0.00
0.00
0.00
0.00
0.00
98.38 1,~: "'''2.67 3.90
20.90 3
,. 26.20 0.78
0.00
0.00
0.00
0.00
0.00
0.00
13.49 32.88 24.45 1013. 5 10.88
1.94 5.05 5.01 19.94 1.57
5.93
0.80
6.92
0.93
0.00
0.00
0.00
0.00
136.05 114.72..~2 156.18 54.99
19.~4 16.~.9 20.72 6.71
0.99
0.13
0.00
0.00
0.00
0.00
0.00
0.00
0.00,
0.00
7.~., .99 481.14 84.32
9.89
1.33
2.97
0.40
0.00
0.00
0.00
0.00
0.00
0.00
17.94 21.64 422.13 5.35
2.68 2.90 55.93 0.71
0.00
0.00
0.00
0.00
0.00
0.00
32.36 96.38 27.67
4.74 12.77 3.67
0.00
0.00
0.00
0.00
0.00
0.00
27.95
3.11
6.23
0.90
6.52
0.84
0.00
0.00
0.00
0.00
40.92
4.55
1.00
0.11
1.29
0.19
0.00
0.00
27.95
3.11
0.00
0.00
,
59
W
C
::J
.&>.
0
I-
i=
23.68
3.40
58
..
::i
57
1
. '1.49
65.01
c'
5.93
0.00
0.00
3.99
0.44
1.00
0.1t
94.95
14.20
56
0.00 142.65
0.00 21.33
55
54 -+----+---...
53
-14
-12
-10
-8
·6
-4
•
-2
0
LONGITUDE
2
4
•
6
8
10
12
14
-
----------------
• Mature autumn spawners tt986).
56
..
"
55
."
...-
.,
"
54
,-
53
-14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
Mature autumn spa,vners (1987).
63
ABUNDANC
tt:
BIOMASS
-~
62
,.
"!'t
2.00
0.51
0.00
0.00
0.00
0.00
0.00
0.00
104.00 0.00
2&.75 0.00
0.00
0.00
0.00
0.00
0.00
0.00
10.00 13.00
2.&3 3.42
&.00
1.68
0.00
0.00
18.00 16.00
4.73 4.20
0.00
0.00
180.12123.68 265. 7 21.77 15.10 10.38 12.21
29.78 18.30 43.79 3.30 2.90 2.00 1.37
0.00
0.00
0.00
0.00
61
3.43
0.51
6.00
1.58
60
,
59
W
~
N
C
::J
to-
i=
5.6&
1.09
10.38 10.93
2.00 1.23
3.86
0.43
0.00.
0.00
9.44
1.82
8.49
1.63
3.86
0.43
2.57
0.29
8.36
0.94
101.83 77.60 20.76 101.90 0.00
15.75 12.87 3.99 19.61 0.00
58
;
:J
57
"
56 -
37.92 215.31
4.26 24.17
35.41 97.19 35.86
5.29 18.70 6.90
0.00
0.00
6.43
0.72
2.57
0.29
99.62
11.18
50.94 43.94
7.68 6.&2
0.12
0.01
0.09
0.01
0.54
0.04
0.32
0.03
1.18
0.09
0.22
0.02
108.89 0.08
16.41 0.01
0.39
0.03
0.99
0.08
3.41
0.27
0.09
0.01
0.03
0.00
192.94 0.01
29.07 0.00
2.34
0.18
0.89
0.07
0.72
0.06
0.03
0.00
1.21
0.09
0.04
0.00
0.88
0.07
0.24
0.02
0.00
0.00
0.00
0.00
0.00
0.00
4.12
0.31
2.59
0.19
0.28
0.02
0.40
0.03
0.12
0.01
2.75
0.22
0.95
0.07
0.78
0.06
1.07
0.08
12.24
0.96
4.66
0.36
7.77
0.61
2.29
0.17
5.90
0.44
8.92
0.70
4.64
0.36
7.43
0.58
1.&9
0.13
0.00
0.00
0.00
0.00
0.00
0.00
3.14
0.25
1.96
0.15
55
54
0.48
0.04
53
-14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
.'
6
8
10
12
14
-.lo . ",:..., ......__
.~-
----------------
...
. ..
e Mature anturnn spa,vners ~88).
63
ABUNOANC
BIOMASS
~~
-"'S
.
62
..
~
~".
".
211.00 69.00
37.97 12.42
\.
227.41 129.27 590.00 23.00 574.00
52.80 30.02 106.18 4.14 103.30
61
45.80 390.80 ~.89 193.00 0.00
10.63 87.
. .14 34.73 0.00
..
15.3~ 0
59
O.~.
O.Q
UJ
C
J::o,
UJ
30.71
4.13
6.94
0.80
5.94
0.80
32.69
4.40
44.91 218.54 32.69
6.88 33.47 4.40
0.00
0.00
,.6.52
302.27 7.08
46.30 1.08
100.4~~
:J
Ij::
~6.01
65.02
15.10
60
~
3.00 39.17 118.06 176.33 0.00
6.00 18.0B 23.72 0.00
26.91
4.12
0.00
0.00
'
...
55
28.30 28.30
2.95 2.95
0.00
0.00
28.30 28.30
2.95 2.95
33.51 37.35 38.20
3.75 4.18 4.27.
36.92 126.77 61.19
4.13 14.18 6.84
56
"
7.78 101.04101.04 0.00
1.19 13.59 13.59 0.00
64.16 274.40 0.00
8.30 36.91 0.00
57
'.
9.56
1.28
24.05 51.03 2B.30
3.40 7.58 2.95
. '26.50
"
:)
13.18
1.92
, i
lf',
58
70.00 32.00
12.60 6.76
28.30 28.30 28.30
2.95 2.95 2.95
1.70
0.19
4.17
0.52
4.17
0.52
4.17
0.52
6.50
0.38
0.00
0.00
37.77 105.48 47.99
4.22 11.80 6.37
4.17
0.52
4.17
0.52
4.17
0.52
6.50
0.38
4.00
0.31
0.00
0.00
0.00
0.00
1.00
0.10
6.74
0.53
6.74
0.53
0.37
0.04
4.67
0.36
0.00
0.00
15.59
1.58
1.00
0.10
6.74
0.53
6.74
0.53
0.37
0.04
0.00
0.00
8.76
0.92
1.00
0.10
8.00
0.73
8.00
0.73
0.37
0.04
0.00
0.00
19.10
2.00
54
13.87
1.45
53
-14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
-----~--------------------------------------------------------
lVlature autumn spa\vners (1989).
63
ABUNDANC
BIOMASS
~I
62
~
18.00
4.60
\.
0.00
0.00
0.00
0.00
0.00
0.00
141.53 49.66 426.00 196.00 59.07 51.09
24.56 8.62 108.76 50.04 13.03 11.27
61
1037.0 472.61 f.0.40 413.08 99.54 158.85 94.99
179.95 82. . 4.68 81.92 19.74 35.05 20.96
"'1l2.09 456.87103.52 0.00 164.44 52.30:
e. 31.60 90.61 20.53 0.00 36.28 8.35 '.'
6.62 212.71
1.15 36.91 1
60
298.79 88.56 168. 2 146.50 223.02 46.95 17.48 39.95 11.21
51.85 15.37 29.15 25.42 40.79 8.59 3.72 8.50 1.79
•
144.67389.84 179.01 43.28 61.17 28.09
23.31 67.64 32.74 7.92 13.01 5.98
59
t
W
C
::J
t-
i=
96.09 171.01 2.16
12.66 28.96 0.34
57.59 33.17 40.87 24.37
9.19 6.06 7.46 3.52
7.04
1.28
43.89 22.67
8.01 3.27
0.00
0.00
5.70
1.04
21.94 41.24 21.06
3.04 5.72 2.92
0.00
0.00
0.58
0.00 106.17352.72 0.00
0.00 14.72 48.92 0.00
1.75
0.24
0.00
0.00
61.15 61.92 95.74 34.55
8.06 9.92 15.28 5.51
58
.
«
-J
0.83
0.14
57
"
56
55
8.77
1.27
2.90
1.42
0.19
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.05
0.00
0.35
0.03
0.35
0.03
1.48
0.20
1.48
0.20
0.00
0.00
0.05
0.00
0.35
0.03
0.35
0.03
1.48
0.20
1.48
0.20
0.00
0.00
0.00
0.00
0.00
0.00
54
0.00
0.00
29.49
4.26
0.90
0.08
0.00
0.00
0.05
0.00
0.00
0.00
0.20
0.01
0.20
0.01
0.00
0.00
0.20
0.01
0.20
0.01
0.00
0.00
53
-14
-12
-10
-8
-6
-4
-2
0
2
4
6
8
10
12
14
LONGITUDE
•
.
.- "
.
e
e
Mature autumn spa,vners (1990).
63
..'.
.(
;
,-..
.
~:..
2.79
0.61
~~;
-'If6.
62
~
25.03
6.51
310.16726.06127.66156.69128.03
67.92 159.70 28.08 34.50 28.19
\.
105.00 289.77 775.93 88.76 237.02 398.67 6.71'
22.99 63.46 170.67 19.52 62.19 87.79 1.14,11
61
ta;;
5.99 167.88 200.62 fO.84 123.00 40.00 128.75239.65 18.9!:"
1.31 36.76 43J~ 3.69 25.50 8.29 25.38 47.23 3.01.'
'.~
60
84.90 381.34
.~11~1.75 207.00110.00 4.70 207.70 247.96~ .
18.59 83.51 1 .. ,.28.85 42.92 22.81 0.93 40.94 39.29.
'.'1.
+----+---I-----+----+----t----+-::-#---+---t---_+_~
198.40101.52 128. 5 79.69 140.00 76.00 22.33 100.13 35.40 .~.
43.45 22.23 21.32 17.45 26.37 14.32 4.21 19.31 6.26
,
81.55,~
16.5S.·
5
59 -
22._:. ;.82 162.40226.09664.33161.25 148.52 70.35 17.30
3,l . '2.57 23.41 46.02 116.86 28.36 23.74 11.17 2.66
UJ
0
...
i=
:J
+-Vl
9
77.42 29.55 98.00 148.00 118.01 123.99 31.2t.
11.91 6.20 18.46 27.88 22.24 23.81 4.67
28.11 101.04 220.31 152.12 2.88
4.04 19.32 38.75 26.76 0.46
58
88.45 35.01
15.56 6.16
9.13
1.46
15.63
2.50
9.77
1.22
7.87
0.96
67.38 167.38 107.65 107.65 0.00
22.86 22.86 9.72 9.72 0.00
0.00
0.00
2.38
0.27
2.38
0.27
f
<t:
..J
57 -
56 +----f-----'I-----+-----LJR,
.
y.
,....•
55
0.54
0.06
0.16
0.02
0.15
0.02
47.63 47.63
4.36 4.36
6.15
0.62
5.15
0.62
0.65
0.07
0.54
0.06
0.15
0.02
0.15
0.02
47.63 47.63
4.36 4.36
5.15
0.62
5.15
0.62
0.65
0.07
8.70 198.20 198.20 0.00
1.28 28.42 28.42 0.00
0.00
0.00
3.35
0.33
3.35
0.33
0.00
0.00
8.70 198.20 198.20 0.00
1.28 28.42 28.42 0.00
0.00
0.00
3.35
0.33
3.35
0.33
0.00
0.00
0.50
0.06
.'
.:.
57.33 18.27
9.12 2.78
54
0.50
0.06
0.00
0.00
0.00
0.00
0.05
0.00
0.05
0.00
0.00
0.00
0.00
0.00
0.50
0.06
0.00
0.00
0.00
0.00
0.05
0.00
0.05
0.00
0.00
0.00
0.00
0.00-
,
53
-I----+---+--'O"-
-14
-12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
Nlature auturnn
53
sp~nvners
(1991).
-+----+----f~
-14
-12
-10
-8
-6
-4
-2
0
2
4
LONGITUDE
•
6
8
10
12
14
e Mature autumn spa,vners (1992).
e
63
+----+----t----+-----t----+----+---_+_---t----+----+-"7'"'."~t:c,~~_:__~___:_'_:__::_--:-''._::_:_:_:__-__:t_
ABUNDANCE
~
BIOMASS
62
.
_t_--_+_---t__---+--=~_+_---t__--_t_--_+_---+-----+--_.<=
t;v
5.88
1.31
0.00
0.00
7.40 154.40 15.10 69.23 90.11
2.07 43.16 4.22 17.24 20.11
6.14
1.32
0.00
0.00
\.
0.00
0.00
32.70 56.40
9.14 15.77
5.91
1.65
61 -+-----+----t----I__---+----f----+---------jl__--_+_---+--q;;
0.00
0.00
98.40 143.70 ~7.20 182.40 74.88 65.62 41.32
27.51 40. 0'1)9.10 50.99 20.07 14.64 8.86
.,g4.00 42.80 11.95 55.83
•. 23.48 11.96 3.20 12.46
0.58
0.13
60 -+----+----+---------j---+----+----+---:--?}----+----t----+---1
46.89 57.26 102.73 21.63 54.47
12.33 15.06 27.02 5.69 12.14
59
0.00
0.00
92.84 208.38 75.24
24.42 54.81 14.99
C
I-
~
-J
58
37.50 111.26113.73 28.70 79.33 139.58111.66
8.36 24.79 25.34 6.40 17.70 29.94 23.95
-+-----+----+---I----+---~-+--~:~-+------'I-----+----+---~
w
:J
96.67 10.72
25.43 .2.82
19.70 51.90 84.23
5.51 13.98 18.79
108.66 187.10307.82 265.24 0.60
21.22 38.59 64.74 56.01 0.11
62.14 231.30 280.32 129.11 125.66 35.21
12.18 45.33 54.94 25.30 24.63 6.62
0.00
0.00
-+----+---t----+--~'.
19.53
3.89
53.57 61.69
10.08 11.61
4~8
0.91
78.26 50.48
15.60 10.06
57 -+-----+----+---j-----..,r----'"<
27.62 252.70 21.40
5.41 49.53 4.19
1.20
0.23
0.00
0.00
18.06
3.40
0.00
0.00
5.56
1.03
0.30
0.06
6.32
1.19
8.62
1.39
15.46
2.08
10.50 92.67 256.99
1.94 17.14 47.52
56 -+-----+----+---I-------lW(}
17.92
3.31
0.00
0.00
35.83
6.63
66.10
12.22
7.41
1.37
0.00
0.00
14.21
2.63
55
-+-----+----+---Tt':~'"-
54
_t_---+---z:,
53
-+-----+-----t---:::.c.....-....;.;...:..;-:-~~--t----"~.;.;..;..;..--'--'T-'-.......;..;....;...;........r-z.--_+_--__+_---1L.-.l..~~.....,;...~~.:..:..;...~;.;..;.;.......:..:..;..~~-:..;..;;.__+_
-14
-12
-10
19.15
3.54
2.47
0.46
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
-
-
-
-
--------------------------------------------------------------
-
Mature autumn spawners (1993).
63
_+_----f---
ABUNDANCE
~
BIOMASS
62
00
~
~
0.00
0.00
0.00
0.00
1.91
0.52
0.38
0.09
14.03
3.27
2.27
0.53
0.00
0.00:
99.28 107.33 241.78 49.73
27.12 29.32 66.06 13.32
3.79
0.88
1.52
0.35
O.OQ·
0.00
7.08
1.93
.,
61
-+----+----t-------Ir-----t----+-----+------l---+----+_~
60
-+----+----t-------If------t----+-----+--:--?il-----l----t-----+--~
59
LU
C
::J
::. I-
,.
.
-+-----+-----+------l----==""~;._+---+_--_+__--__+---+_--+_-,;;;
0.00
0.00
0.00
0.00
84.17 71.92~7.49 34.81
22.99 19. Ir: 4.83 9.51
24.63 476.71
6.65 125.67
9.93
2.4.6
1.52
0.3S:
•
74.09 187.04 30.94
16.80 30.79 4.32
14.89 W!4.tl4 132.81 168.90 39.60 202.68 126.58 28.91
3.6 _......« ;2 17.91 46.15 10.82 34.n 20.84 4.04
7.611.06
-t----+---t__---t-----+--~t__-t_Sb'__+_--_+---+___--___t_--
4.1'&'11 .23
-' 0.66
1
0.00
0.00
58
--+-------I----+----+_~
57
- t - - - - - - + - - - - J - - - - - t - -....----"ö---fT>
..J
76.46 210.54 238.50 146.39 109.58 48.19
11.55 52.61 54.48 25.04 18.04 6.73
4.31
0.60
30.72 46.30 129.13 22.68
7.30 26.05 4.08
2.83
0.29
26.58 14.15
2.69 1.43
1.78 139.15 4.95
0.18 22.26 0.79
0.00
0.00
0.21
0.02
4.84
0.49
10.75
1.09
33.65
5.38
9.52
1.52
1.24
0.13
0.73
0.07
0.30
0.03
6.37
0.64
3.50
0.31
6.09 608.93 77.10
0.62 92.23 11.68
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
48.50 216.84 37.29
4.93 32.84 5.65
0.00
0.00
0.00
0.00
0.55
0.07
0.00
0.00
0.00
0.00
14.37 69.05
2.90 13.93
1.12
0.11
492.55 :83.6.17
53.15
10.72
56 - + - - - - + - - - j - - - - - - + - - - - ' L l C
55 -t----+----+----,'r';:-
54
10.23 23.12
2.38 5.38
'11;0.58 88.66 27.12 36.39 27.67
.43.87 24.22 7.04 8.47 6.44
171.21 30.66 25.02 14.58 135.94 7"5.
198.51 6.12
42.45 7.60 6.20 3.98 36.32 10.13 54.23 1.67
0.00
0.00
4.23
1.01
0.95
0.11
1.73
0.20
0.00
0.00
1.96
0.23
0.00
0.00
0.00
0.00
-t----+---
53
-14
-12
-10
-8
-6
-4
-2
0
2
4
LONGITUDE
•
6
8
10
12
14
e
1\tlature autumn spawners (1994).
63
,
.,
.').
--
•
ABUNDANCE
~"';
.,
,
BIOM ASS
62
':t
;..
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
29.12
7.73
0.00
0.00
0.00
0.00
0.00
0.00
37.80
10.03
0.00
0.00
0.00
0.00
0.00
0.00
40.31 21.29 209.94 36.89 S'O.
28.95
11.98 3.91 55.71 5.69 12.38 7.68
0.00
0.00
0.00
0.00
0.00
0.00
9.87
1.54
25.85 191.93 68.68 178.67~~ 129.80 46.32
7.68 ~5.25 12·Vl 47. ;:\.... 4 20.02 12.29
8.74
2.32
0.00
0.00
0.00
0.00
7.31
1.02
6.74.
0.94
4'~'1I 0.74 164.29 202.82 0.00
1
. 1.44 25.34 53.82 0.00
0.00
0.00
4.18
0.82
27.15
5.31
3.82
0.55
177.12 33.75 41.13 13.15
26.60 8.96 10.92 3.14
0.00
0.00
0.00
0.00
92.74
20.58
50.07 34.65 24.02 13.55
6.88 4.79 5.18 3.24
0.00
0.00
62.97
13.98
5.79
0.88
11.60 25.61 39.15 492.47 0.00
1.59 3.54 8.02 117.57 0.00
0.00
0.00
3.97
0.60
0.00
0.00
60
0.00
0.00
5.37
1.60
59
0.00
0.00
W
C
•
0::
43.43
12.90
::>
!:: 58
\0
~
..J
57
-.
56
133.97 54,28
29.57 11.98
0.00
0.00
53.31 163.75 29.26
14.15 43.45 7.76
0.00
0.00
61
J::,.
1.04
0.28
25.16
6.68
••
~i
29.14 146.62 335
5.87 32.36 67.
9.45 743.25 20\0.23 35.
2.08 164.03 48.38 7~
10.75
1.48
0.00
0.00
32.36 5.88
6.52 1.18
0.00
0.00
0.00
0.00
5.14
0.68
0.00
0.00
7.35
1.62
rP"
22.
4.49
0.74
0.10
3.80
0.52
"
~
....
55
,',
"
54
53 - + - - - - f - - - - f - . . = . l
-14
·12
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
6
8
10
12
14
l\tlature autunl" spawners (1995).
63 - f - - - - + - - - - + - - - - f - - - - - ! - - - - f - - - - + - - - f - - - - f - - - - f - - - - - f - -........
ABUNDANCE
BIOMASS
45.72
8.8,8
59
w
Ul
o
175.93.;1;~
260.26
38.30
+----+-----:.~f----+----..:.j..:.-----.,;.-+1-8.'j~.49 387.57
3~ . 5.08
68.38
C
::J
I-
34.~~0
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
2.97
0.79
0.00
0.00
0.00
0.00
9.99
2.68
0.00
0.00
1.04
0.19
89.44 0.00 0.00 7.95
17.30 0.00 0.00 1.48
99.93 15.00 39.59 11.29
17.68 2.27 5.94 2.10
0.00.
0.00
0.00
0.00
2.17
0.33
33.81
6.29
76.37 73.17 69.29 59.46
12.07 11.39 10.70 8.92
7.19
1.34
6.09
1.13
42.96 21.18
6.44 3.20
1.36
0.21
0.56
0.10
1.12 128.63 24.61
0.21 14.91 3.30
0.00 364.20 2~.83 0.0
0.00 73.42 38.49 0';
0.00
0.00
15.01
2.85
3.21
0.61
0.00
0.00
0.00
0.00
0.00
0.00
0.00
0.00
9.41
1.35
2.46
0.32
7.57
0.99
29.47 723.37 269.83 560.44 13.94
3.82 108.71 42.63 88.55 2.13
17.68
3.56
t= 58 -+----j---4----+
27.20 2q.55 0.00
O;~'"
<C
5.48
..J
0.00 223.52 0.0
0.00 45.06 O.
43.05
.
o.
;-"
57
83.12
12.54
184.46
33.82
0.00
0.00
2.07
0.42
75.70
9.66
"'0.0
0.0
56 - + - - - - + - - - - + - - - . f - - - - "
0.00 118.29 40.20
0.00 24.17 8.21
0.00
0.00
13.32 17.68
1.74 2.31
190.51
38.93
55 -t----+----+---,
0.52
0.07
0.00
0.00
0.00
0.00
54 -t-----+----n
0.00
0.00
5.63
0.93
53
-14
-12
0.00
0.00
-10
-8
-6
-4
-2
0
LONGITUDE
2
4
•
6
8
10
12
14
... ....... .., ........... ..
~
,"
~'
-
--------------------------------------
e'l\tlature autumn spawners (1996).
e
63 - t - - - - t - - - - t - - - t - - - - - t - - - - t - - - - t - - - t - - - - - t - - - - t - - - - - f A6UNDANCE
610MASS
62 - t - - - - t - - - - t - - - t - - = . J I l t t - - t - - - - t - - - - t - - - t - - - - - t - - - - t - - - . : ;
~
~
61
60
22.13 20.44
5.83 5.38
0.00
0.00
0.00 504.20 267.28 ff9.86 252.31 22.97 70.70
0.00 '.112.78 55. iJ7.48 54.14 4.15 18.61
3.25
0.86
469.89 115.36 0.98
105.11 25.80 0.22
73.08 451.71 1
9.23 18.06 0.00 40.19 5.09
16.35 99.87 2
.20.49 3.47 0.00 10.58, 1.34
-t------if-----t-----+---0-.0-0-+3-5-.3-,-22-7-.6-0+-9-68-.-51-,-'-9-.6-2r)7'."0-2-6-0-.9-l0r O-.-00--0-.0-0+7-.-,0 ----i---'1
f2
18.52 0.00
0.00 7.40 47.68 216.64 26.76 25.06 58.36 0.00 0.00 1.87 4.88 0.00
.
•
59
0.00
0.00
LU
C
VI·
~
I-
~
..J
28.42 97.32 46.38 180.82 169.7~~ 190.04 263.56 4.64 0.00 0.00 0.00 0.00,
5.95 20.39 ,6.85 26,99 24.~.•-a;.a7 39.25 58.95 1.04 0.00 0.00 0.00 0.00
44.43' 8.49 0.00 0.~""~.-'1+4-38-.2-5-2-'-3.-9-l4-17-6-.6-9-,,-.5-7+0-.0-0----i---;
0.00 0.00
9.31 1.25 0.00 0
. 2.09 70.28 38.42 29.98 1.92 0.00 0.00 0.00
58
0.00
0.00
99.52 21.65 49.97
14.80 3.76 8.28
38.57 14~.40 26.09
7.25 26.76 4.""
3.71
0.55
0.00
0.00
0.00
0.00
0.00
0.00
28.06 105.33 0.00
3.n 14.38 0.00
0.00
0.00
0.00
0.00
20.36 94.68 1Ö.
3.83 17.79 1.
57
0.00
0.00
fl1' '.
99.61 54.56 12.
18.72 10.25 2.28
0.00 123.53 7.58
0.00 16.09 0.99
o.n
10.06
1.70
0.49
0.05
0.23
0.02
0.22
0.02
7.23
0.78
0.26
0.03
0.01
0.00
6.13
0.66
1.28
0.14
56 - t - - - - t - - - - t - - - t - - - - - L J
0.00
0.00
18.23 108.74
2.37 14.16
0.00
0.00
0.13
0.92
0.16
55 -+----+----1----,
0.00
0.00
0.55
0.14.
0.00
0.00
54 - + - - - - - f - - - ,...
4.85
1.20
5.05
0.05
0.01
0.00
0.00
35.16 26.95
2.95 2.56
62.13
8.34
29.73 139.78 2~.67
5.59 26.26 38.08 0';
22.51
2.09
1.2~:
53 -t----+----+-....::.J
-14
·12
·10
·8
·6
·4
-2
0
LONGITUDE
2
4
6
8
10
12
14
APPENDIXC
A COMPARISON OF INDICES OF HERRING STOCK (INDEPENDENT
01" THE ASSESS:VIENT MODEL)
John Simmonds, SOAEFD Marine Laboratory, PO Box 101
Victoria Road, Aberdeen - 13 December 1996
Introduction
In order to get some information on relative trends in the abundance of hcrring dcscribcd by different indices a
simple analysis of acoustic and IBTS surveys was carried out. Both surveys provide age disaggregated indices of
abundance which can be used to give indices of relative cohort strength recruiting to the adult stock. These
indices can then be used as an indication of the relative performance of the two series over the period compared.
The indices values used are those reported in the Herring Assessment WO Repart (Anon, 1996).
Methods
The year class strength measured by acoustic and IBTS surveys can be seen for years 1984 to 1996 in Figures 1
and 2 for the two surveys. The IBTS exhibits much steeper selection (faster apparent reduction of numbers at
age) than the acoustic survey, but far each survey it may be assumed that the relative abundance at age in each
year is an indication of year class strength. In order for the comparison to be independent of any changes in
mortality between year class er between surveys only fully comparable year class strengths have been included in
the analysis, thus for a comparison of the acoustic survey and the IBTS only the abundancc of 2, 3, 4 and 5 ring
fish have been included in thc analysis. For some treatments the 5 ring class is not used as this is a +ring dass in
thc IBTS.
•
The relative abundance of each cohort was estimated from the survey time series by two methods:
1. The abundance at each age \veighted by their relative abundance in the survey; this assurnes the errors are
dependant on the abundance, then the relative cohort abundance is:
N 2(y+1) + N 3(y+2) + N 4(y+3) + N 5(y+4)
y=1996 a=5
L LN /LY
ay
y=1984 a=2
where N2(y+ 1) is the number of 2 ring fish in year y+ 1 and Sy is the number of years of data. Where the 5
ring observations arc excluded these are removed from bOlh top and baltom of the equation. These series are
described as the weighted series.
2. Equal weight far each age, to remove thc effects of selection which is kno\'m to be different in each index, in
this case the relative cohort abundance is:
N 5(y+4)
y=1996
+
L,N 5y
y=1984
N 4(y+3)
y=1996
L,N 4y
y=1984
+ N 3(y+2) +
y=1996
L,N 3y
y=1984
N 2(y+l)
y=1996
L,N 2y
y=1984
This series is described as the normalised series.
Results
The relative cohort abundances derived from IBTS and acoustic surveys are shown in Figure 3. The subSlantial
peak in the IBTS in the weighted series is due to a single survey and year class observation; this can be seen in
E:\PFaPGHERS97\REP.DOC
19/03/97
52
•
---
--- -
-
------------1
Figure 1 as an unusually high value in the 1987 cohort occurring at age 2 ring in the 1988 survey. This cohort is
not seen to the same extent at I ring in 1987 or as 3 ring in 1989. The normalised series are less sensitive to this
value.
The ratio of the lETS cohort indices and the acoustic survey cohort indices (from Fig. 3) is shown in Figure 4 for
the 1 ring in 1983 to 1993. For cIarity a three year running mean has been incIuded on the plot. The mean
recruitment ratio from the acoustic to lETS surveys for the first four years of the period can be compared with
the last four years of the period (these two ratios are independent of one another). The normalised series shows a
relative increase of 35% in the acoustic index, the weighted index shows a 5% increase. It is possible to obtain
an indication of the precision of the relative cohort indices using the three estimates of each cohort in the
normalised series. The two normalised series + 1 standard deviation can be seen in Figure 5. This analysis
suggests that the lang term relative performance change in these two surveys is much laee than that indicated by
comparison of the assessment and the acoustic indices, and that the difference between these surveys may not be
significant. It is also possible to estimate the mean CV for the acoustic and lETS series. These are estimated as:
=
=
CV for relative lETS normalised series
CV for relative acoustic normalised series
•
•
E:\PFaPGHERS97\REP.DOC
19/03/97
53
0.43
0.25
Figure 1. Cohort Abundance North Sea Herring trom IBTS
surveys (years are surveys)
5,000 r - - - - - - - - - - - - - - - - - - - - - - ,
4,000
3,000
\
2,000
•
Figure 2. Cohort Abundance North Sea Herring trom
Acoustic Surveys (years are surveys)
5,000,----------------------,
4,000
3,000
2,000
1,000
54
•
- - - - - - - -
Figure 3. Estimates of relative cohort strength (refered to 1
ring year) from 2 to 4 ring and 2 to 5 year age classes,
weighted by survey abundance and weighted byequally.
18TS
2-5 norm
18TS
2-4 norm
18TS
2-5 abs
18TS
2-4 abs
ACOUST
2-5 norm
ACOUST
2-4 norm
ACOUST
2-5abs
1984 1985 1986 1987 1988 1989 1990 1991 1992 199
•
ACOUST
2-4 abs
Figure 4. Ratio of Relative Acoustic and IBTS Estimates of
cohort strength (refered to 1 ring year )from 3 surveys
ldotted)and smoothed (solid) using ages 2 to 4.
2.b
r---......,
2-5 norm
2-4 norm
/
()
2-5 abs
........."!""........
1.5
2-4 abs
... ...
~
2-5 norm
1
2-4 norm
0.5
2-5 abs
o
2-4 abs
1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 199
55
Figure 5 Estimates of relative cohort strength ( refered to 1
ring year) from succesive surveys .:!:.1 standard deviation
IBTS 2-4 years Acoustic surveys 2-4 years
5 , - - - - - - - - - - - - - - - - - - - - - - , ,------,
IBTS
4
Coefficients of Variation
IBTS
cv=0.43
Acoust cv=O.25
3
Acoust
IBTS+1sd
IBTS-1sd
A+1sd
2
A-1sd
OL...l.----'------L.--I.--......I...-_-l.-_-l.-_-l.-_...L.-_...L.-_.LJ
1983 1984 1985 1986 1987 1988 1989 1990 1991
•
1992 1993
•
C-s
56
APPENDIXD
PERCEPTIONS OF NORTH SEA HERRING STOCK DYNAMICS AND SURVEY
VARIAßILlTY THAT ARE ROBUST TO CATCH l\IISREPORTING
KR Patterson, SOAEFD Marine Laboratory
PO Box 101, Victoria Road, Aberdeen
1.
•
This note explores what change in perception of trends in stock size, fishing mortality, landings from thc North
Sea herring stock occurs when thc assumption that catches are estimated without bias is relaxed to the assumption
that catches are unknown. Variances with which the various survey indices estimate stock size and age-structure
are estimated, together with a simple measure of time-series correlation in the errors in spawning biomass
estimates.
2.
•
Introduction
The Model Described
The model used here is a derivative of models described by Patterson (1996) and Anonymous (1996). Familiarity
with those" documents will be assumed and only abrief dcscription of thc present application is provided herc.
The model trcats estimates of fish abundancc separately from thc estimates of age-structure, and ignores
interdependcnce in the two sorts of information. This allows biological sampling of landings for age-structurc to
bc uscd in the model while not using information on total catches, which is here considered a potential source of
bias and is not used. Similarly, acoustic survey estimates of stock abundance (which are principally determined
by echo-integration) can bc treated separately from thc cstimates of thc agc-structure of the stock (which are
derived from the accompanying trawl sampling). For consistency, the IBTS survey has been trcated in a similar
way, and overall herring abundancc from that survey has been trcated as bcing measurcd separately from thc agecomposition of thc fish population. This treatment is consistent with thc usual assumption that thc greatest
source of variability in surveys arises from variable catchabiltiy of the gear.
2.1
Assumptions
The model has been structured with the following assumptions:
•
•
•
•
•
•
•
•
2.2
Catch data are assumed to be unreliable and are not used.
Biological sampling of catches for age-structure information is assumed to be unaffected by misreporting of
catches.
Fish aged less than two years and more than eight years old are excluded from the animalysis.
Random sampling for age is assumed in both catch and in survcys.
Instead of attempting to cstimate annual fishing mortalities, a polynomial trend in fishing mortality with time
is fitted. This is a simple approach to smoothing, which is used because models of this type perform
extremely poorly for estimating annual fishing mortalities or landings.
Landings reported between 1975 and 1981 are assumed known preciscly. Imposing this constraint allows a
rescaling of the missing catch model for comparison with a conventional model.
Selcction is assumed to bc uniform from ages 2 to 8.
Notation
Dcfining thc following variables:
a,y - subscripts describing year and age
N - Population abundancc
W- weight of fish at agc and year
P - proportions of fish at agc and ycar in thc catchcs or in thc survcys
U - index of abundancc of spawning fish
E:\PFaPGHERS97\REP.DOC
19/03/97
57
Q - proportionality parameters ("catchabilities") relating the indices of abunance of the larvae and acoustic
surveys to the fitted populations.
F - fishing mortality
fi - parameters of the fishing mortality trend in time, i =I,j
T2 _ effective sampie size when sampling for proportions at age (see Haist et al., 1993 for a detailed description
of this parameter).
2
.
(j - survey vanance
npc, nu • number of observations of proportions in the catches and in the surveys respectively.
S- estimates of the variances of the proportions of fish by age in the catches, calculated on the assumption of
random sampling from a multinomial distribution. Sis calculated as PC.,y(I-PC.,y)+ 0.01 (Haist et al., 1993).
Cl - as S' for sampling for ages in acoustic surveys.
2.3
Struetural assumptions
Fishing mortality and abundance estimates from historic VPA have been assumed to be known precisely up to
1981. Thereafter, fishing mortality is structured as
(I)
i.e., a simple polynomial-time model. Additional terms fi are added to the model until adding additional
parameters results in no significant improvement to the model fit at the 0.05 probability level (terms are not
included if the log-likelihood increases by less than 1.96).
•
The usual exponential mortality and catch equations are assumed to hold and are used as the structural model.
As absolute estimates of stock size cannot be calculated when catches are not known, all abundance estimates are
scaled to VPA-estimated values for the period between 1975 and 1981.
Spawning stock biomass is simply calculated as
SSB y =
2.4
La N a,y Wa,yOa,y maturity ogive
(2)
Information on age-structure
The model is fitted by finding a maximum in the joint log-likelihood term for age-structured information and for
information about the indices of abundance of spawning biomass. Omitting subscripts showing the source of
information (surveys, catch sampling, etc) the 10g-like1ihood component is:
(3)
where S is given by
(4)
and T 2 is estimated iteratively as
(5)
For surveys, the proportions P are proportions of fish by age in the sampIes. For commercial
catches, they are proportions of fish by age in the samp1ed catches, with corresponding conventional structural
assumptions.
E:\PFOPGHERS97\REP.DOC
19/03/97
58
•
2.5
Information on abundance of spawning biomass
These are treated conventionally assuming a simple proportionality of survey index to the spawning stock size
from the structural model, and lognormal observation error:
1" (
--.L.J
2
Y
2
In(2ITa)+
In(QB/ U y )2
a
2
J
(6)
with variances iteratively recalculated as
•
0"'=Ly( ln(QB~~ U y)'
J
(7)
Separate estimates of 2 for each source of age-structured information were calculated, and separate estimates of
of Q for each estimate of SSB were also calculated.
2 and
3.
•
Data
The information used to estimate stock parameters was:
•
•
•
•
•
•
Estimates of the proportion of total international catch by number in the adult (2+ stock), by year (1975 to
1996).
Acoustic survey estimates of the abundance of spawning fish by year (1986 to 1996)
Acoustic survey estimates of the proportion of fish by age in the adult stock (1986 to 1996)
IBTS survey estimates of the biomass of fish (1983 to 1996)
IBTS survey estimates of proportion of fish by age in the adult stock (1983 to 1996)
The multiplicative larval abundance index (MLAI) from International Herring Larvae surveys
These data are given in Anon (1996), except for the 1996 acoustic larval survey estimate (Simmonds, pers.
comm., 1997)
4.
Model Fits and Parameter Estimates
Key population parameter estimates are given in Figure 1. This shows generally similar trends to those estimated
in the conventional assessment model, except that:
•
•
•
•
Recruitment between 1985 and 1990 is estimated as being much higher
Catches in the period 1987 to 1991 are estiamted as being substantially higher than reported catches, almost
reaching double reported levels from 1987 to 1990. However, as noted in Patterson (1996), the precision
with which catches or fishing mortality can be estiamted with this type ofmodel is very low.
The model indicates that the spawning biomass of hcrring in the late 1980s may have reached much higher
levels than indicated in the conventional model. According to estimates of accuracy in Patterson (1996),
estimates of stock size from this type of model are expected to be quite good.
Details of the fuH model fit are given in the Appendix Tables.
Estimates of the variance of survey spawning biomass estimates are:
Acoustic SSB
0.32
Probability of no positive
autocorrelation in errors (*)
0.02
MLAI
0.26
0.06
IBTS SSB
0.21
0.79
Survey
Variance «j2)
(*) Runs test, Draper and Smlth (1981, p157)
E:\PFC\PGHERS97\REP.DOC
19/03/97
59
Estimates of the T 2 parameter, indicating the effective sampIe size for estimating the proportions of fish by age
are:
T2
Source of information
Catch sampling
0.0164
Acoustic surveys
0.0022
IBTS age-structure
0.232
The above model fit and associated estimates are obtained by using aII available information, and making the
assumption that the observation erorrs are independent. That however may not be the case, to the extent that
different surveys may have different time-trends in residuals, which is indicated by the runs-test diagnostics. As
a further data explotation exercise, the model fits were repeated but either fitting only to a) catch and acoustic
survey data, b) catch and larval survey data, or c) catch and IBTS data. The same parameters were estimated,
which allows a simple exploration of the dependence of perceptions of survey precision on making prior
assumptions that each of the surveys in turn is "correct" with respect to the other two.
The perception that the age-structure of acoustic surveys is measured accurately contrasts with the perception that
the accuracy of the spawning biomass estimate from the same survey is poor. To explore the consistency of the
surveys further, the model was fitted to the commercial catch sampling proportions and to the acoustie survey
age-structure alone. The resulting fit agrees weil with the remaining sources of information, with the exception
of the acoustic survey SSB estimate (Fig. 6).
Data series
Fitted to acoustic
SSB+ ages
0.01644
Fitted to
IBTS SSß+ages
0.01549 -
Fitted to acoustic
a.ges
0.0164
Fitted to
MLAI
0.00215
0.00925
0.00215
0.00953
IBTS age-structure (r)
0.235
0.197
0.366
0.471
Acoustic SSB (er)
0.0866
0.375
I.I57
0.257
IBTS SSB (a 2 )
0.898
0.196
0.366
0.250
1.44
0.376
0.336
0.257
Catch sampling (r)
Acoustic survey age
structure (T2)
MLAI (er)
5.
•
•
0.01517
Discussion
The following inferences could be drawn from the parameter estimates above:
1. In terms of measures of variability in abundance estimates, the IBTS abundance index performs best, the
acoustic survey performs worst, amI the performance of the MLAI index is intermediate.
2. In terms of measures of between-year correlation in errors in abundance, which may be more important in
terms of providing advice for management purposes, the IBTS survey is unlikely to have error
correlation(P=0.97), the Acoustic survey is very likely to have correlated errors (P=0.02) and the MLAI index
is somewhat less likely to have correlated errors (P = 0.06).
3. In terms of estimating the age-structure of either the catch or the stock, the acoustic survey has the largest
effective sampie size, the IBTS has the smallest effective sampIes size, and the sampling of commercial
catches is intermediate.
Overall the model suggests that the most rcIiable sourcc of information are the acoustic survcy cstimatcs of agcstructure and the IBTS spawning biomass estimates.
Such inferences are of course predicated on the assumptions made in Section 2.1, and rcly on ignoring process
errors (eg changes in selection pattern, changes in natural mortality,etc).
E:\PFarGIIERS97\REP.DOC
19/03/97
60
•
6.
References
Anonymous. 1996. Herring assessment working group for the area South of 62 N. lCES CM Assess: 10.
Draper, N. and Smith, H. 1981. Applied regression analysis (second edition). lohn Wiley and Sons, 709pp.
Patterson, K.R. 1996. Assessing fish stocks when catches are misreported: model, simulation tests and
application to cod, haddock and whiting in the ICES area. ICES CM 19961D:7.
•
•
E:\PFOPGHERS97\REP.DOC
19/03/97
61
-
-
---------------------------------------------------------------~
Appendix Table 1. North Sea Herring. Elements of the Missing Catch structural model
Relative Populauon abundance (Populations In 1975· VPA estlmatesj
2
3
4
5
6
7
8
;,um
J.~
Relative
1975
839
360
198
71
23
11
4
1976
826
148
70
43
15
5
2
1977
2BB
177
31
16
10
4
1
1978
251
103
70
14
7
4
2
1979
425
177
80
60
12
6
4
1506
1.00
121C
0.80
527
0.35
451
0.3C
764
0.51
1980
4B6
296
136
66
51
10
5
1054
0.70
1981
1406
274
185
94
47
35
. 7
1982
1383
747
161
120
61
31
23
1983
2674
831
496
118
88
45
22
1984
5472
1558
535
353
84
63
32
1985
6135
3092
973
369
243
58
43
1986
6376
3362
1673
651
247
163
39
1987
14209
3389
1975
1216
423
160
106
1988
19335
7325
1931
1243
765
266
101
1989
11644
9667
4047
1179
759
467
163
1990
5936
5646
5180
2397
698
45C
277
1991
4333
2792
2935
2976
1377
401
258
1993
3&67
2046
977
760
883
896
414
1994
5710
1702
970
512
398
463
469
15121
10.04
1992
4624
1999
1407
1635
1658
767
223
12314
8.18
2049
1.36
2526
1.68
4274
2.84
8096
538
10914
7.25
12711
844
21478
1426
30&66
20.56
27926
1854
20584
1367
9944
6.60
10224
6.79
1981
179
51
41
22
13
11
2
1982
176
1984
675
294
120
85
23
19
10
1985
633
587
219
90
66
17
14
1988
1670
1122
416
307
208
75
31
1989
1243
1595
855
285
205
132
48
1226
8.26
1627
1096
1986
66&
634
419
161
66
47
12
201e
1354
1987
1197
605
434
297
115
45
33
319
2.15
36
29
17
9
7
413
2.78
1983
340
155
111
28
24
13
7
67&
457
2726
1836
3829
2579
4363
2939
1990
636
1026
111e
577
186
127
82
3745
25.22
1991
415
541
637
705
354
111
76
1992
328
369
303
384
438
213
68
1993
217
242
209
182
242
261
130
1994
538
264
201
115
105
134
146
2839
19.12
2102
14.16
1483
999
1995
4792
2376
763
493
26C
203
235
9142
6.07
1996
3598
1934
106C
386
243
128
100
7449
495
1996
284
362
241
88
64
37
32
1975-80
1504
10.13
1995
460
446
176
116
71
60
75
1404
946
1107
7.46
148
100
1996
048
026
0.14
0.05
0.03
0.02
0.01
Mean
Relative Spawnlng Biomass In the Population
Age
2
3
4
5
6
7
8
o
I
'-I
0'1
N
;,um
J.~
Relative
1975
107
67
44
17
6
3
1
246
1.66
1976
118
28
16
10
4
1
1
178
1.20
1977
37
33
7
4
3
1
C
1978
32
19
16
3
2
1
0
1979
54
33
18
14
3
2
1
85
0.57
74
050
126
085
1980
62
55
30
16
14
3
2
183
1.23
1979
056
023
011
0.06
002
0.01
0.00
1980
046
028
013
006
0.05
0.01
0.00
1981
069
0.13
009
DOS
0.02
0.02
0.00
1982
055
030
006
005
0.02
0.01
0.01
1983
063
0.19
012
003
002
0.01
0.01
1984
068
019
007
0.04
0.01
0.01
000
1985
0.56
028
009
003
0.02
0.01
000
1986
0.50
026
0.15
005
002
0.01
o.oe
1987
066
016
009
006
002
0.01
0.00
1988
062
024
006
0.04
0.02
0.01
000
1989
042
035
0.14
004
0.03
002
001
1990
029
027
025
0.12
0.03
0.02
0.01
1991
029
0.18
0.19
0.20
009
0.03
0.02
1992
0.38
0.16
0.11
0.13
0.13
0.06
0.02
1993
0.40
021
0.10
0.08
0.09
0.09
0.04
1994
0.56
0.17
009
0.05
004
0.05
0.05
1995
0.52
0.26
0.09
0.05
0.03
0.02
0.03
1981
0
0.333
0.333
1982
1
0.550
0.210
1983
2
0481
0240
1984
3
0424
0271
1985
4
0.273
0.301
1986
5
0436
0332
1987
6
0.315
0.363
1988
7
0.226
0.393
1989
8
0.168
0424
1990
9
0.261
0.454
1991
10
0.247
0.485
1992
11
0.180
0516
1993
12
0.189
0.546
1994
13
0.164
0.577
1995
14
0.164
0.607
140
Proportions In the Population by Age
Age
2
3
4
5
6
7
8
1975
0.56
024
0.13
0.05
0.02
0.01
0.00
1976
0.77
0.12
006
004
0.01
0.00
000
1977
0.55
034
006
003
0.02
0.01
0.00
1978
056
023
016
003
0.02
0.01
000
FIshing Mortallty Parameterisatlon
fears trom start
WG Mortahty esl.
F,tted F
1975
1976
1977
1978
1979
1980
1.434
1434
1.355
1.355
0.728
0.728
0.048
0.048
0061
0.061
0.272
0.272
e.
•
Appendix Table 1 (contd.) North Sea Herring. Elements of the 'MCM' structural model
Calches In Number
Ages
2
3
4
5
6
7
8
SuM
'<eI. 1975
1975
571
254
145
52
17
8
3
1051
1.00
1976
613
102
50
30
11
4
2
812
0.77
1977
131
84
15
8
5
2
1
246
0.23
1978
10
4
3
1
0
0
0
19
002
1979
22
9
5
3
1
0
0
40
0.04
1980
101
64
31
16
12
2
1
227
0.22
1981
347
71
50
25
13
10
2
517
0.49
1982
227
128
29
22
11
6
4
427
041
1983
497
161
101
24
18
9
5
815
0.78
1984
1129
337
121
80
19
14
7
1707
162
lS85
1390
733
242
92
60
14
11
2542
2.42
1966
1569
866
505
176
67
44
10
3237
308
lS67
3768
940
573
353
123
47
31
5834
555
1968
5484
2173
599
386
238
83
31
8994
856
1969
3512
3048
1335
369
251
154
54
8743
832
1990
1894
1883
1807
836
244
157
97
6917
6.58
1991
1473
980
1078
1093
506
147
95
5371
5.11
1992
1630
736
542
829
638
295
86
4::>57
434
1993
1462
787
393
306
355
360
167
3831
3.65
1994
2193
683
406
215
167
194
197
4054
386
1995
1913
990
341
215
113
88
102
3763
3.58
1977
16
15
3
2
1
0
0
38
0.62
1978
1
1
1
0
0
0
0
3
0.05
1979
3
2
1
1
0
0
0
6
0.10
1980
13
11
7
4
3
1
0
38
0.62
1981
41
10
9
5
3
3
1
72
1.16
1962
27
19
5
5
3
1
1
61
0.99
1983
59
24
18
5
4
2
1
114
1.84
1984
133
50
22
17
5
4
2
233
3.77
1965
178
120
47
19
13
4
3
384
6.22
1966
190
132
92
36
15
10
3
478
7.75
1987
373
140
103
74
29
12
9
740
11.98
1966
609
315
104
76
51
20
6
1183
19.15
1969
404
466
231
81
58
38
14
1292
20.92
1990
216
281
320
161
56
37
24
1095
17.72
1991
191
163
198
222
110
35
25
943
15.27
1992
168
129
102
130
142
70
21
763
12.36
1993
168
114
74
62
81
88
43
631
10.21
1994
285
109
74
46
40
49
54
656
10.63
1995
318
192
70
47
27
22
27
702
11.36
1977
053
0.34
006
0.03
002
0.01
1978
054
023
0.17
0.03
0.02
0.01
000
1979
054
0.24
0.11
0.08
0.02
001
001
1980
044
028
0.14
0.07
0.05
001
001
1981
Otl7
0.14
0.10
0.05
0.02
0.02
000
1982
1983
081
0.20
0.12
0.03
002
0.01
0.01
1984
066
020
007
0.05
001
0.01
000
1985
0.55
0.29
0.10
0.04
002
0.01
000
1986
0.48
0.27
0.16
0.05
0.02
0.01
000
1987
065
0.16
0.10
0.06
0.02
001
0.01
1988
081
024
007
004
0.03
001
000
1989
040
035
0.15
004
0.03
0.02
0.01
1990
027
0.27
026
0.12
0.04
002
001
1991
0.27
0.18
0.20
0.20
0.09
0.03
0.02
1992
038
0.16
0.12
0.14
0.14
006
0.D2
1993
038
0.21
0.10
008
0.09
0.09
0.04
1994
0.54
0.17
0.10
0.05
0.04
005
0.05
1995
0.51
0.28
0.09
0.06
0.03
0.02
003
Calches In Welght
Ages
2
3
4
5
6
7
8
Cl
I
0'1
W
co
Sum
Rel. 1980·1975 m
1975
72
45
31
13
4
2
1
167
2.71
1976
77
18
11
7
3
1
0
117
1.90
Mean 1975-1980
Proportion In calches
Agos
2
3
4
5
6
7
8
1975
0.54
024
0.14
0.05
0.02
0.01
0.00
1976
076
0.13
006
0.04
0.01
000
0.00
ooe
0~3
0.30
007
0.05
0.03
001
0.01
61.7671
.------------------------
-
Recruitment
Landings
;
25 . . . - - - - - - - - - - - - - - - - ,
CIl
:E
25
B
111
B
4i
g: 15
.~
~..,
1V~
-..,
&!
c'"
111'- 10
E
~
5
Si 10
~
.:l
15
_t--
1Il'-
:cc::
20
~
t'll
20
111 0
...
5
t)
111
~
o-l-!!:II\,g.m.lItill:j!=-----;-----+----l
1975
1980
1985
1990
0
1975
1995
Year
1985
1990
1995
Year
Biomass
Fishing Mortality
1.6...---------------,
:e
30
1.4
.'"
25
111
20
B
u. 1.2
11 1
.~ 0
,..----------__.r----..,
E
g: 15
111.-
~ 0.8
~
~
0.6
w 0.4
III
III
E
O+--....,.;;!!!-I------+----+------i
1980
1985
1990
•
10
ffi
0.2
1975
1980
1995
5
0 8;I~iJ!II~~-_+_--_+_--.-J
1975
1980
1985
1990
1995 ;
Year
Year
Figure 1. Summary of missing catch model fit compared with the assessment model fit
given in Anon. (1996).
•
D-9
64
Figure 2. Residual-time plots for the three surveys giving information about Spawning
Stock Biomass. Log residuals for the full model fit are plotted.
0-10
65
IBTS SSB Index
~
~
~
'tl
.!
'§,
'äj
~
:;)
2.3
1.9
1.5
1.1
0.7
0.3
-0.1
-O.!3
-0.9
-1.3
-1.7
l--------+-------+----=--\---+~t_----+--_l
5
1980
-2.1
-2.5 - ' - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - '
Year
Acoustic SSB
~
~
Cll
~
•
2.3
1.9
1.5
1.1
0.7
0.3
1-------+------~--~
•.....,____::A"~~+:::.~
... ...........
._-_+_-_1
~
~
~
.! -0.1
'§, -O.~H 175
1980
1985....
-0.9
~ -1.3
:;) -1.7
-2.1
_2.5.1-.
1990
~
-
'Qj
----1
Year
MLAI
~
~
2.3
1.9
1.5
Cll
1.1
0.7
'tl
0.3
0::
.! -0.1
'§, -o.~
J:-----+-t-------t----~__i---~-_+-----_j
1
-0.9
~ -1.3
:;) -1,7
-2.1
-2.5 - ' - - - - - - - - - - - - - - - - - - - - - - - - - - - " - - - - - - '
7
'jjj
Year
Figure 3. Residual-time plots for the three sUNeys giving information about Spawning
Stock Biomass. Log residuals when fitting the population model to the acoustic sUNey
and to the catches alone,
0-11
66
•
I
IBTS SSB Index
1.8
~
:::l
"tl
.~
~
:E
-r-------------------------------,
1.4
1
0.6
0.2
-o.~
/------_+_---~:___/_1I--~~f--_\_-_+:;~\;_.....----'I_\_--1
5
.~ -0.6
~
:>
.
1980
-1
-1.4
.1.8 -'--
--J
Year
Acoustic SSB
1.8
~
-r-----------------------------,
1.4
:::l
1
.~
0.6
~
:E
-o.~
"tl
0.2
/-------+-------t------~"+------+---1
5
.~ -0.6
~
:>
1980
1985
-1
-1.4
-1.8
-1..-
----'
Year
MLAI
1.8
•
~
:::l
"tl
.~
Cl::
"tl
:E
1
0.6
0.2
-o.~
.~ -0.6
~ -1
:>
-r------------------------------,
1.4
-1.4
-1.8
I------li------+----\---+--~+--"....__+_--_____.~--t
2
-1..-
.....
Year
Figure 4. Residual-time plots for the three surveys giving information about Spawning
Stock Biomass. Log residuals when fitting the population model to the IBTS and to the
catch sampling data.
0-12
67
I
IBTS SSB Index
1.8,------------------------------,
~
'tl
'in
~
1.4
1
0.6
0.2 I - - - - - - - + - - - - - - - - J ' l - - " '.......-+-~.___M~_'\:__...-..--r'\-_l
-O·i 5
1980
.~ -0.6
'tl
~
~
::J
r
-1
-1.4
-1.8..1------------------------------'
Year
Acoustic SSB
~
1.8
1.4
1
'tl
.~
-r------------------------------,
0.6
~ 0.2 \--
~
-O'i
5
.~ -0.6
~
::J
•
-+
1980
-+-
--,>"-+
-+-_-j
1985
-1
·1.4
_1.8..1---------
---J
Year
MLAI
1.8.-----------------------------,
(ij
1.4
::3
1
.~
0.6
'tl
a:::
'tl
0.2
'Qj
-0.6
~ -O.~
~
::J
1-------1~--\:__:AH~~~4,__-+~~~-~. ...,.._:At___J_c---~
2
-1
-1.4
-1.8
-L-
--'
Year
Figure 5. Residual-time plots for the three surveys giving information about Spawning
Stock Biomass, Log residuals when fitting the population model to the MLAI larval
index and to the catches alone.
0-13
68
•
.
IBTS SSB Index
2.5
2
~
1.5
~
1
0.5
t
-O.~
~
"i
.~
-r-----------------------------,
0+---------11------:;.---~....,_--+__'<_-_+_~"_:___,;.c----+-----=-~
5
1980
-1
:5
-1.5
-2
-2.5 - ' - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - '
Year
Acoustic SSB
•
~
~
2.5
2
1.5
~
1
0.5
t
-o.a
"i
-.----------------------------~---,
O+--------t--------t-------:~'----------+--_.j
.~
-1
:5
-1.5
-2
-2.5
5
1980
1985
-1-
---1
Year
MLAI
2.5
-r----------------------------,
2
~
1.5
~
1
0.5
t
-o.a
~
"i
.~
:5
O+----------;f+-----"L---+--~~._-_+_--~'--""'*-_jk-+--__j
2
-1
-1.5
-2
-2.5 - ' - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -.....
Year
Figure 6. Residual-time plots for the three surveys giving information about Spawning
Stock Biomass. Log residuals when fitting the population model to the catch sampling
and acoustic survey age-structure information.
0-14
69
•
APPENDIXE
WORKING DOCUl\IENT TO TIIE PLANNING GROUP FOR IIERRING SURVEYS • FEßRUARY
1997. EFFECTS OF REDUCED SAl\1PLING EFFORT ON AßUNDANCE AND PRODUCTION
ESTIMATES FROl\l NORTII SEA IIERRING LARVAE SURVEYS
J Gröger and D Schnack*
Institut für Ostseeforschung, An der Jägerbäk 2, D-18069 Rostock, Gennany
*Institut für Meereskunde, Düstembrooker Weg 20, D-24105 Kiel, Gennany
Introduction
Duc to a substantial declinc in ship timc and sampling effort allocated to thc Herring Larvac Survcys since the
end of thc 80s, it has been qucstioned, whether thesc surveys can still providc abundancc and production indices
(LAI and LPE) comparablc to thosc of prcvious years and sufficiently reliablc for the usc as measure of stock
sizc. Sincc 1992/3 a multiplicativc model was uscd to fill in missing values (Patterson and Bcveridge, 1994,
1995a) and for thc period 1994/95 no traditional indcx was calculatcd, but thc multiplicativc model approach was
used to analysc overall trends in thc larval data series (Patterson and Bcveridgc, 1995b). This method assurnes
that larval production in each area and time unit defined for thc traditional sampling schedulc is a certain constant
proportion of thc total. In vicw of thc urgent demand for reliablc stock size cstimates, thc Herring Assessment
Working Group requestcd to cvaluatc thc validity of assumptions in thc methods uscd and to define the minimum
sampling effort required and possiblc survcy strategics that could bc achieved considering given restrictions in
ship time.
•
•
The Planning Group for Herring Survcys rccognised that it will not bc possible to perfonn further detailcd
analyses and testing of survcy strategies at Aberdeen and, thus, suggested to transfer thc entire data set of the
herring larvac surveys, presently located in Aberdeen (Scotland), to Kiel (Germany) and to update this set
regularly. Maintenancc of thc data bank and standard analyscs for assessment purposes, however, were still to bc
carricd out at Abcrdcen for thc timc being.
This working document summarises thc present status of implementation of thc data bank and standard analyses
at Kiel and of the requested specific analyses.
Implementation of Data ßank and Standard Analyscs of IIcrring Larvac Sun'cys at Kiel (Gcrmany)
A complete set of thc herring larvae data has first been established in Rostock. The structure of thc data bank has
been c1arified and thc calculation routincs for standard analyses of abundancc indiccs havc been established
according to thc definitions given in working group reports (Anon, 1985, 1986), manuals and summary
presentations (Heath, 1992), and personal information obtained from Aberdeen. Thc calculation of both indices
(LAI and LPE), however, include some final problems still to bc solved:
Thc LPE valuc is critically depcnding on zIk estimates, obtained from thc length distribution of larvae. The
sampled larvac populations did not always providc a reasonable basis for calculating these values. Thus, they
havc frequently been estimated as some mean from previous surveys. The way this problem has been handled
seem to differ between years and thc values utilised are difficult to identify. It appears to be essential to define a
standard procedure for estimating the zIk values.
For calculating the LAI there remain somc uncertaintics in the area values used to raise the mean abundance
values per station, and these areas appear to differ from those used in the LPE.
Hoth thcsc problems havc to be solved befoTe the indices can be calculated for individual successive time peTiods
separately, within any of thc standard areas in a way that they are comparable to the combined yearly values
given for these areas. It is planned to transfer the system to Kiel after it has been completely established.
E:\PFOPGIIERS97\REP.DOC
19/03/97
70
•
•
Validity of Assumptions Madc in thc Multiplicath"c Model Approach
It has already been mentioned by Patterson and Beveridge (l995b) that the assurnption for this model approach
may be invalid, as the distribution of herring spawning areas have undergone strong changes. Figure I compares
the proportion of larvae abundance for the four traditional spawning areas in the North Sea. The LAI values for
ShetlandlOrkney are taken as standard and those from Buchan, Central North Sea and Southern North Sea are
given as percentage compared to this standard. It is obvious that there are significant changes in the relative
importance of the individual areas between extended periods: The large percentages for the compared areas in
the 80s and mostly small values in the 70s and 90s up to now, do indicate a long-term trend, which is to some
extend common among two or three areas but different from the SheltlandlOrkney area.
•
Due to this condition the present analyses did not consider the model in more detail but focused on evaluating the
cffect of using only two or three target areas on the relation between abundance or production index and
spawning stock size as derived from VPA. The effect of timing of the surveys has not yet been analysed duc 10
the above mentioned problem in recalculating the indices. The area effect could be studied by the available
indices from lhe given data set.
Effect of Reduced Sampling by Using Targct Areas on tbc Reliability of Abundancc Indices
•
Area specific regressions of abundance indices on SSB and also Figure 1 indicate that it might be reasonable to
use the ShellandlOrkney area and either the Buchan or the Soulhern North Sea or may be these three as target
areas. The results from combining the LAIs and LPEs from these areas (two and three) have been compared to
those from combining a11 areas including VIa north as there is a large scale advection from west of Scolland to
lhe northern Noith Sea areas (Heath, 1992). The combination has been done by simply using the corresponding
sums of the index values.
Figure 2 provides scatter plots for LAI and LPE comparing tolal coverage with the combination of
Shetland/Orkney and Buchan. Including the southern North Sea results in an intermediate situation. The main
question is whether the variance explained by regression is reduced by using the target areas or not and if so,
whether it could be increased by adding complete coverage in certain intcrvals.
Figures 3 and 4 show for LAI and LPE respectively, the effect of an increasing distance bctween complete
covcrage on the slope, the residual variance (standard deviation = RMSE) and the percentage explained variance
(r 2 = RSQ). The RSQ-value is a redefined value for a model excluding the effect of the intercept, which was in
no case statistically significant. Thus, the absolute value of RSQ is higher than from the regular model. The last
values for a 23 years distance includes no complete coverage in the given data series. The counling starts 1972,
but due to missing values the data series includes a total of 18 data point (1976-1993).
•
The main result appears to be, that there is no substantial effect on the explained variance (RSQ) for bolh indices
and both area combinations. The RMSE values for LAI are exceptiona11y high for one year and for 19 years
distance of complete coverage. This is duc to an exceptiona11y high and fairly uncertain value in the total LAI
value for 1990, which is included in the data set only in these two i!1stances. The values for LAI are fairly stable,
whereas the LPE is more sensitive to the specific year included as complete coverage. But again the reduction to
the target areas does not reduce the explained variation of the regression, and inclusion of complete coverage
does not seem to provide any improvement.
The problem of defining the precision of the required inverse prediction of SSB from the larvae indices has not
yet been sufficiently addressed. Some attcmpt, however, has been made to look for trends in the inverse
prediction error by calculating the variance of the predicted SSB on the basis of a method described by Neter et
al. (1985) for utilising calibration regressions. Figure 5 presents, as an example, prediction errors from a very
low and a very high abundance index value for the different steps from complete to only target area coverage.
The absolute values are not meaningful, as extremely low and high values have been used in comparison to see if
this may have an effect on the trend. There may be some slight but not substantial increase in the prediction error
towards the reduction to the target areas in case of LAI. For LPE this is not obvious within the range of variation
of these values.
E:\PFaPGIIERS97\REP.DOC
19/03/97
71
•
Conclusion
The results lead to the general impression that restriction of the sampling to only two target areas, including
ShetlandlOrkney and Buchan or southern North Sea does not reduce the reliability of stock size estimates. As the
southern stock component is managed separately, this area should be included in any case and it is to be expected
that the combination of this area with Shetland/Orkney will provide comparable results to those obtained with the
combinations presented as exarnples in this paper. This should be checked, however. Additional complete
coverage do not seem to improve the results but could be required for other purposes, eg to check for changes in
the system compared to the observed data series. If the problem of estimating reasonable zIk values for the LPE
ca1culations is solved, this method might be favourable, as it is to be expected that only one coverage could be
sufficient in each of only two target areas. The question of timing of the surveys has already been discussed in
some detail by the Working Group on Herring Larvae Surveys (Anon, 1990). But after adequate standardisation
of the ca1culation method, the effect of timing should again be studied for both indices in view of restriction to
the target areas.
The present results do not indicate that estimation of stock size based on the regression with larval abundance
indices for a combination of the two or three target areas mentioned here would be less reliable than base on a
complete area coverage.
References
Anon. 1990. Report of the Working Group on Herring Larvae Surveys South of 62 N. ICES CM 19901H:32
Anon. 1985. Report ofthe Working Group on Herring Larvae Surveys South of62 N. ICES CM 19851H:3
•
Anon. 1986. Report ofthe Working Group on Herring Larvae Surveys South of62 N. ICES CM 19861H:3
Heath, M. 1992. An evaluation and review of the ICES Herring Larvae Surveys in the North Sea and Adjacent
Waters. Contribution to American Fisheries Society Early Life History Section Los Angeles 23-27 June
1991. Proceedings Los Angeles County Natural History Museum.
Neter, J., Wassermann, W. and Kutner, M.H.
Homewood-I1linois.
1985. Applied linear statistical models. Richard D. lrwin,
Patterson, KR. and Beveridge, D. 1994. Report of the Herring Larvae Surveys in 1992/l993. ICES CM
19941H:25
Patterson, KR. and Beveridge, D. 1995a. Report of the Herring Larvae Surveys in the North Sea and Adjacent
Waters in 1993/l994. ICES CM 19951H:22
Patterson, KR. and Beveridge, D. 1995b. Report of the Herring Larvae Surveys in the North Sea and Adjacent
Waters in 1994/l995. ICES CM 19951H:21.
E;\PFaPGHERS97\REP.DOC
19/03/97
72
•
•
...
Area LAis relative to Shetl/Orkn.
--- Buchan
--*--- Centr.NS ..•--- South.NS
140
120
100
Q)
Cl
as
.....
c
80
...0
60
(])
Q)
a.
40
20
0
1970
1975
1985
1980
1990
1995
Year
Figure 1
Trends in herring larvae abundance indices (LAI) für Buchan, Central Nürth Sea
and Süuthern North Sea as percentages üf the value für the Shetland/Orkney
area.
•
E-6
73
LAI vs SSB
•
7 0 . . . - - - - - - - - - - - - - - - - - - -1
60
.
j •
50
--g
Total
...................... __ _-
*
.
.
•...............
- -
SH+B 1
t·
·nn
n
•••
40
(l)
Cf)
:J
a:s
I-
---
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .* "
••
•
............................................................................................................•.........................................................................
30
20
•
1O
•
.*='""
*
•................. ~
~
"'~
*
o
0,25
,*
•
.
0,5
0,75
SSB
(Millionen)
1
1,25
•
LPE vs SSB
4000..--------------
3000 ..... /
•
Total
SH+B
-,
/
.
•
w
a..
~
2000
.
'*
•
1 000
•
........................................•................. ..- ..............................•....*
•
*
*
•
~
.
*
*
0-+--'-----,-----..----------.----.---_
o
Figure 2
0,25
0,5
0,75
SSB
(Millionen)
1
1,25
Scatterplots of herring larvae abundance index (LAI) and production estimate
(LPE) versus spawning stock biomass (SSB) based on herring larvae survey data
from all North Sea spawning areas including subarea VIa north (Total)
compared to combined results from Shetland/Orkney and Buchan area only
(SH+B).
74
•
LAI: Effect of area reduction (Sh + B)
•
10000.------------------------.100,0
•
co
j
(f)
(f)
(/)
>
80,0
8000
<!
-<
.....J
10
........
o
-?fl.
6000
T"""
i<
Q)
0
(f)
er:
0...
o
(j)
"U
c
40,0
4000
ro
•
w
(f)
\
:2:
er:
Siope
~-.-.-~~---.to.-.-..-.-~
2000 ---,----,----,---,.--.-.-...,--.,-r---=i=----r---,-----.---.----r---,.-,..-...,--..---"r----.--r-----r-+ 20,0
o
5
10
15
20
No of years between comp!. surveys
+.
•
LAI: Effect of area reduction (Sh+B+D)
10000..--------------------------.100,0
........
ca
(f)
(f)
CJ)
>
<!
80,0 .
8000
.....J
L ()
•
........
<
0
T"""
6000
60,0
i<
Q)
?fl.
-.
0
(f)
er:
00
(f)
"U
c
ro
w
(f)
:2:
er:
4000
40,0
~e... ..
:--A-A--A-.-.---...-----.--A-....-..
A---.l..-.Il.-...
2000 -I-.----,--..,--r---,--r----.---,---,.---,.-,-...,--.,-.----r--r-r---.----r---,.-,-,..-.--+-20, 0
5
10
15
20
o
No of years between comp!. su rveys
Figure 3
Regression of LAI on SSB: Effects of reduction in sampling effort by increasing
the number of years between complete surveys (total. area coverage) and
sampling in between in two or three target areas only. SH+B: Shetland/Orkney
and Buchan as target areas. Sh+B+D: Shetland/Orkney, Buchan and Downs
(southern North Sea) as target areas. Regression parameters: Siope, Standard
error of Residuals (Root Mean Square Error = RMSE), and R2 (RSQ given as
percentage).
75
LPE: Effect of area reduction (Sh +8)
900-,-,--------------------,100,0
...-...
co
Cf)
Cf)
(/)
>
w
a...
-
-l
~
50J- ·
......................................................···················60,0
&~
o
Cf)
CI:
~ 300l~:~----100
•
20,0
I
o
5
.
10
15
20
•
No of years between compl. surveys
LPE: Effect of area reduction (Sh+B+D)
m
900....----------------------.-100,0
RSO
Cf)
Cf)
(/)
>
w 700
V\~
\._
.....................................
.
........................................................................ 80,0
a...
-l
l {)
<
~ 500
......................................................···················60,0
<D
""Q.
Cf)
CI:
o
Cf)
"'0
c
~
w
Cf)
~
.... 40,0
300
Siope
CI:
1 00 -f-'----"-r---l----.----r----r-r-'--'---.--.--..-r-l---r---.---,-r-,--,--,--,-t-2 0 ,0
o
5
10
15
20
No of years between compl. surveys
Figure 4
0
Regression of LPE on SSB: Effects of reduction in sampling effort by increasing
the number of years between complete surveys (total area coverage) and
sampling in between in two or three target areas only. SH+B: Shetland/Orkney
and Buchan as target areas. Sh+B+D: Shetland/Orkney, Buchan and Downs
(southern North Sea) as target areas. Regression parameters: Siope, Standard
error of Residuals (Root Mean Square Error = RMSE), and R2 (RSQ given as
percentage).
76
•
LAI: Pred.Err. vs Reduction {Sh+B+D}
•
50..,.'------------------'?~r.------,
·.
··· ...
.'
'
.)
''
Max LAI
«
.....J
40 -
------....... .
c
:2:
~,'
~t" .
<;t
""'"...
'*.
~~
,~
.. *.*. "'*'".,'
I
~."""
""',
*
*
"
'
*' i.
. ,''''~
~ ~
'
\~
.*-.*
€.'''' .•...•••....\ ..•• ;~.....•..••.......•••...
,,"
;t[
I
< 30
0
...-
-l<
•
~
•
...:
....
w
-0
....<1>
20
.
a...
1 0 +-...---~,-----,--r---,------,---,----,----.-.---..,--r-r---r----.--.---.--.---.----.-.---..,---j
o
•
10
15
20
5
No of years between comp!. surveys
LPE: Pred.Err. vs Reduction {Sh+B+D}
50...----------------------r100
LU
a... 40
.....J
C
90
Max LPE
.
........
80
. ...
~H.
••••••
....
~
:2:
.....J
I
I
~,* I
\. T
<;t
70
o
~60
...-
-l<
m
10I
<
o
...-
-l<
...:
....
Min LPE
w
-g....
><
:2:
·········\/················~50
< 30
w
a...
20
40
30
a...
...:
....
W
V
....<1>
a...
20
1 0 -+---'--"'---'----"'---'----'----'----'-----'--'----'-...--.--r---I--.-.---.--.---.----.-.---'-+- 1 0
o
5
10
15
20
No of years between comp!. surveys
Figure 5
Inverse prediction error according to Neter et al. (1985) for SSB, estimated from
extreme low (1) and high (61277) values of LAI and LPE. Effect of reduction in
sampling effort by increasing the number of years between complete surveys
(total area coverage) and sampling in between in three target areas only:
Shetland/Orkney, Buchan and Downs area.
77
•
•
"
APPENDIXG
COMMUNICATION INFORMATION FOR RESEARCH VESSELS
Telephone
Vesse1
Mobile
GO Sars
jBridge: +47 9455 6811
lMess: +4794505071
.............................}jGSM: +4791193383
Scotia
1+44836385975
I
Telefax
Satellite
Mobile
I
Telex
Radio call sign
Satellite
~+871 3257 15010
1::::+4794549900
1:,'+8713257 15012
:+5814257 15010
1+871 3257 15011
!+581 3257 15014
.
.
~
.
················1············
········..··············
!+871 1440552
l+44 836385975*
1+871 1440561
l
1+872 1440552!
1+872 1440561!
!
~
i
1
LLZG
.
.
GOWS
~
. ·..····..··r···.
.· . · · ·..·. · · · · · ·. · . · ·..···~+s·7Tii'2"jii7"·· ···· ·r····· · ·. · · · ·. · · ·············!+87i·iii·3ii"i ·······..··..1+s"8"i··ü"2'3iif···..··· T.· 'DB"FR..·..·..· ·
~.~~!.8..~~.~ .L
L
L
L
h:?~}. ~~~} . ~~.?.?.9.
L
.
~~it~;
Irridens
!
5~~~· ·..· . ·.. ··ls~·i~~·ti~t~+4·5""3·0~i"8 . 6·8·64···..·i+s·7Ti6Toios..·..· ···
:Bridge: +45 3020 0363
E:\PFQPGHERS97\REP.DOC
19/03/97
l
!
j
r..·
:
··..·
···..···..·····..··
1
1+87i..j·6'i"020f..··..·..· t+5S·i··i6"i..020S
~
:
·..····.. t · ·oxl31(..····..
~
•
•
.. •
APPENDIXH
PLANNING GROUP FOR HERRING SURVEY - ABERDEEN 24·28 FEBRUARY 1997
Name
felephone No
lRichard Aukland
........................................................
lMartin Bailey
IFacsimile No
IE-mail
lauklandr@marlab.ac.uk
!
ibaileymc@marlab.ac.uk
j+44 1224295539
1+44 1224295511
~un
~
[+44 1224295539
!+44 1224295511
..
~~~·i·F~;~~d~~·····················t~4··1·224··295·403················1~44··i·2·24·2·95·5·i·l················lf~~~~~d·~~~~@·~~i~b~~~~~k···············
:
;
:
~b·~~h~d·GÖ·;;~····················r~49··40··38·90·S2·32··············1~49·4o·3·89Ö·52·64·············T·····
.
~·ii~··Häk~~~~~~····················t~6··523·1··87·i6·····················t:46·s2"3"i··3·97·7·····················I~~·h~k~·~~~~·~·@·i~:~;····························
...........................................................~
. . . . . . . . . . . . . . . . . . .n
••• n
••u
•• n n
···..···..·····
····
~+49 40 3890 5232
k:ornelius Hammer
nn.nn
~
nnn
nn
·..······1···..·..····
····
!
··•
~+49 40 3890 5264
~
nn.n
n
.
~nilshammer@aol.com
~ ..................................................................................................• • • • • •
.. n
Kenneth Patterson
[+441224295507
l+44 1224295511
ipattersonkr@marlab.ac.uk
.........................n
I
i
1
,1
~~~~··R;id·····_························~44··i224··295363················i:44··i·224··;·95s·i·l················I~;idd~@·~·~i~b:·~~:~k··········..··············
lDietrich Schnack
I
.......................................................................... Jo
60
~ Simmonds (Chairman) 1+44 1224 295366
.
Idschnack@ifm.uni-kiel.de
'
~+44 1224295511
.
lsimmondsej@marlab.ac.uk
~:~~~:~~~~:~~~~:~~:::::::::::::::t~~::~:~~~::~~~~:::::::::::::::::::::1~~~:~~:~~:~:~~:~:::::::::::::::::::::H:~~:~~~:.:~:~~~~:~::::::::::::::::::::::::::::::::::::::
~~.~~.~. :~.~~~~.~
.J~:=?.~.~~~ . ~=~.~
.J~~:=.:.~.::. ~.~~.:
..i~.~.~~.~~.~~~:~.~
.
~7
lEise Torstensen
E:\PFaPGHERS97\REP.DOC
1+473705 9000
19103/97
3705 9001
80
lelse.torstensen@imr.no
,.