ICES WKADR REPORT 2008 Reports/Expert... · 2013. 2. 4. · ICES WKADR REPORT 2008 3 . 1.3 Workshop structure a nd working procedure Participants are listed in the following Section

ICES WKADR REPORT 2008 ICES ADVISORY COMMITTEE

ICES CM 2009/ACOM:57

REF. SCICOM, AFWG, NWWG, PGCCDBS, PGRS

Report of the Workshop on Age Determina-tion of Redfish (WKADR)

2 - 5 September 2008

Nanaimo, Canada

International Council for the Exploration of the Sea Conseil International pour l’Exploration de la Mer

H. C. Andersens Boulevard 44–46 DK-1553 Copenhagen V Denmark Telephone (+45) 33 38 67 00 Telefax (+45) 33 93 42 15 www.ices.dk [email protected]

Recommended format for purposes of citation:

ICES. 2009. Report of the Workshop on Age Determination of Redfish (WKADR), 2 - 5 September 2008, Nanaimo, Canada. . 68 pp.

For permission to reproduce material from this publication, please apply to the Gen-eral Secretary.

The document is a report 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.

© 2009 International Council for the Exploration of the Sea

ICES WKADR REPORT 2008 i

Conten ts

Executive summary ........................................................................................ 1

1 Overview ............................................................................................... 2

1.1 Introduction ................................................................................................................ 2

1.2 Terms of Reference .................................................................................................... 2

1.3 Workshop structure and working procedure ..................................................... 3

1.4 List of participants..................................................................................................... 3

2 Information on age determinations, otolith exchanges and validation work since the 2006 workshop .................................................................. 4

2.1 Otolith exchanges 2007-2008 ................................................................................... 6

2.1.1 Demersal S. mentella from East Greenland ............................................. 6 2.1.2 Pelagic S. mentella from the Irminger Sea and adjacent

waters.............................................................................................................. 9 2.1.3 S. marinus from Iceland .............................................................................11 2.1.4 S. marinus from Iceland, 1990 yearclass.................................................13 2.1.5 Pelagic S. mentella from the Irminger Sea and adjacent

waters, Russian samples ...........................................................................15 2.1.6 Pelagic S. mentella from the Norwegian Sea, Russian

samples .........................................................................................................19 2.1.7 S. marinus from Flemish Cap ...................................................................21 2.1.8 S. mentella from Flemish Cap ...................................................................24 2.1.9 S. fasciatus from Flemish Cap...................................................................26 2.1.10 Pelagic S. mentella from the Norwegian Sea .........................................28 2.1.11 Juvenile Sebastes sp. from the Flemish Cap ..........................................32

3 Sources of age determination error in terms of bias and precision, interpretational differences between readers, ageing criteria ..................... 33

3.1 Results for otolith set A .......................................................................................... 36

3.2 Results for otoliths set B ......................................................................................... 39

3.3 Conclusions............................................................................................................... 41

4 Most appropriate otolith preparation method for age determination of redfish and stock-specific guidelines for ageing........................................ 43

5 Sex-specific growth information and differences in growth estimates between otolith readers and preparation methods ..................................... 44

6 Progress in the implementation of quality control and quality assurance (QC/QA) in redfish ageing labs ................................................ 46

7 Publication of the results of the 2006 and 2008 workshops and the recent otolith exchanges (2007–2008) in the ICES CRR serie ........................ 48

8 Recommendations ................................................................................. 48

9 Literature cited...................................................................................... 49

ii ICES WKADR REPORT 2008

Annex 1 : List of participants ........................................................................ 51

Annex 2 : Agenda ........................................................................................ 54

Annex 3 : Glossary for age determination of redfish (updated from ICES 1996) .................................................................................................... 56

Annex 4 : Protocol for handling and age determination of North Atlantic redfish otoliths ..................................................................................... 58

ICES WKADR REPORT 2008 1

Executive summary

The WKADR held in Nanaimo/Canada 2-5 Sep 2008 was a follow-up of the 2006 workshop with a specific focus to analyze the results of the otolith exchanges carried out in 2007-2008 and to identify the sources of error in age determination. In addition, improvements in age reading methodology and quality assurance/quality control of redfish age reading were discussed based on working examples. The workshop showed significant progress in the correspondence between readers,

The workshop showed significant progress in the correspondence between readers compared with previous workshop results. Also there was a clear improvement in agreement between the exchange exercise and the readings during the workshop af-ter discussions, especially when using only clear-pattern otoliths and comparing the most experienced readers. Good quality otoliths with good readability showed only small differences among readers and demonstrated that in general there are no dif-ferences in age reading criteria among readers. This fact highlighted the importance of using a confidence index for each reading. The workshop encourages to continue the implementation of QA/QC in the different laboratories involved in age reading of redfish. For stock assessment and regular precision monitoring, a confidence index is proposed.

Only otoliths with good or fairly good confidence index should be used for compari-sons and discussion on ageing criteria among readers. However, it remained unclear if only those otoliths should be used for production ageing, especially for stock as-sessment purposes, as these clear-pattern otoliths may correspond to given ages or cohorts and hence may bias the age-length key and the stock age composition if only those otoliths are used. It was also clear that several of the observed differences are considerably reduced when using a high-quality binocular equipment, prepared for large magnification and with high quality light especially for older otoliths (>30 yrs). Standardization of the equipment, and its quality, promotes reproducibility between readers and agencies.

Some differences for certain criteria were detected in the position of the first few an-nuli, the identification of several checks, especially during the first 3-4 years, and the interpretation of the transition zones. Although discrepancies could be due to spe-cies/stock-specific differences, it was agreed that each country should measure the first three annuli in 10 otoliths with very clear patterns for a number of stocks, and containing several year classes. The workshop recommends to create a reference col-lection within each lab for their specific stocks, and their pictures should be trans-ferred to an image collection to be uploaded on a webpage with sufficient metadata. These differences in interpretation very likely are derived from each reader being fa-miliar with some species or stocks. To overcome this problem, it is recommended to continue with the harmonization of the age reading within and across stocks and spe-cies following the guidelines, and through short-term exchanges among labs. To harmonize age reading within stocks, short-term exchanges should also be conducted among labs determining age of the same or related stocks

The break-and-burn technique was established as preferred otolith preparation me-thod for redfish. Several recommendations are given with regard to the improvement of lab procedures, the continuation of age reading harmonization and small-scale oto-lith exchanges. An updated glossary and protocol for handling and age determina-tion of redfish otoliths is given in the annexes.

2 ICES WKADR REPORT 2008

1 Overview

1.1 Introduction

Redfish (genus Sebastes) are an important fishery resource that is distributed and tar-geted by fisheries throughout the North Atlantic. High quality (accurate/repeatable) and comparable ageing is one of the most important, yet unresolved, issues amongst agencies conducting research and stock assessments. Age determination data are the basis of age-based analytical assessment of the species and stocks under investiga-tion. Controversy continues to exist regarding the most appropriate means of age de-termination, and several attempts have been made to create common criteria (ICES, 1983, 1984, 1991, 1996, 2006). While ICES has definitively agreed on the use of otoliths exclusively for redfish age determination, recent studies have revealed that consider-able discrepancy in ageing criteria still exists. The slow growth and longevity of North Atlantic redfish has made the issue of accurate age determination particularly difficult to resolve. Different countries are using different otolith ageing methods for different redfish. The result is large differences in age per length class in, at least, S. mentella and S. marinus around Iceland and in the Irminger Sea.

Due to these discrepancies, redfish otoliths are seldom routinely aged, and hence age based analytical assessment is normally not conducted for any stock. For the alterna-tive length based or age-length based methods, reliable estimates of growth rates are essential. The previous redfish age reading workshop in 2006 analyzed the sources of age determination error and agreed on initiating an otolith exchange for further anal-ysis and agreement. After the exchange the labs and staff involved are in a better po-sition for a further discussion and agreement in an international context.

1.2 Terms of Reference According to ACOM Resolution 2007/2/ACOM34, the Workshop on Age Determina-tion of Redfish [WKADR] (Co-Chairs: F. Saborido-Rey, Spain; and C. Stransky, Ger-many), was held from 2-5 September 2008 in Nanaimo, Canada, to:

a ) review information on age determinations, otolith exchanges and validation work since the age reading workshop in 2006;

b ) identify sources of age determination error in terms of bias and precision, describe the corresponding interpretational differences between readers and laboratories, and agree on the ageing criteria;

c ) identify the most appropriate otolith preparation method for age determination of redfish and provide stock-specific guidelines for ageing;

d ) collate sex-specific growth information and analyze differences in growth estimates between otolith readers and preparation methods;

e ) investigate the progress in the implementation of quality control and quality assurance (QC/QA) in redfish ageing labs;

f ) prepare the publication of the results of the 2006 and 2008 workshops and the recent otolith exchanges (2007–2008) in the ICES CRR series.

WKADR reports for the attention of ACOM, RMC, NWWG, AFWG, PGRS and PGCCDBS.


1.3 Workshop structure and working procedure Participants are listed in the following Section and Annex 1. The agenda for the Workshop as adopted during the opening session of the meeting is provided in An-nex 2.

The Workshop was divided into five activities or sessions: i) Reviewing information on age determinations, otolith exchanges and validation work (ToR a), ii) Sources of age determination error in terms of bias and precision. Ageing criteria (ToR b), iii) identification of the most appropriate otolith preparation method for age determination of redfish (ToR c), iv) analyze growth differences between sexes, readers and preparation methods (ToR d) and v) investigate the progress in the implementation of quality control and quality assurance (QC/QA) in redfish ageing labs (ToR e).The results and discussions of these sessions are described correspondingly in section 2 to 6 of this report.

Session ii) and iii) took most of the time as a large collection of otoliths was read dur-ing the workshop and a selected number discussed in plenary. Each session, as well as the conclusions and recommendations in relation to the terms of reference, were dis-cussed and reviewed in plenary. The final list of recommendations is included in Section 8 of this report.

1.4 List of participants

Ole Thomas Albert Norway Barbara Campbell Canada Karen Charles Canada Nora Crosby Canada Mariña Fabeiro Spain Darlene Gillespie Canada Barbara Grabowska Poland Joanne Groot Canada Sif Guðmundsdóttir Iceland Lise Heggebakken Norway Mary-Jane Hudson Canada Kristján Kristinsson Iceland Svend Lemvig Norway Diana Little Canada Shayne MacLellan Canada Susan Mahannah Canada Judy McArthur Canada Fran Saborido-Rey Spain (Co-chair) Arne Storaker Norway Christoph Stransky Germany (Co-chair) Margaret Treble Canada Kordian Trella Poland

Participants’ affiliations and e-mail addresses are given in Annex 1.


2 Information on age determinations, otolith exchanges and validation work since the 2006 workshop Since the 2006 workshop (ICES 2006), redfish ageing labs have applied harmonised age reading criteria. In general, labs did not change their established otolith prepara-tion method (break-and-burn, break-and-bake, thin sections). Table 1 gives an up-dated overview on the age reading methods by nation and stock.

Some labs have introduced a 'confidence index of readability' in the frame of the oto-lith exchanges carried out 2007-2008 (see section 2.1), as part of a quality assurance framework (see section 6). Table 2.1 provides an overview of these exchanges.

No further age validation work has been carried out on Atlantic redfish since the last workshop. Two Pacific Sebastes species (S. pinniger and S. alutus), however, have been age-validated by radiometric methods (Andrews et al. 2007, Kastelle et al. 2008).

The otoliths collected and read during 2006-2008 were added to an updated overview table (Table 2).


Table 1 Summary of methodological aspects for Sebastes age determination.

COUNTRY STOCK SPECIES STRUCTURE METHOD # AGE READERS

QC/QA OBSERVATIONS

Germany Irminger Sea S. mentella Otoliths Thin-section

(1) Not implemented Otoliths read occasionally

Germany Iceland/Greenland S. marinus and S. mentella

Otoliths Thin-section

(1) Not implemented Otoliths read occasionally

Iceland Iceland S. marinus and S. mentella

Otoliths Break and burn

1 Not implemented* S. marinus read on a routine basis for stock assessment. S. mentella read only occasionally

Iceland Irminger Sea S. mentella Otoliths Break and burn

1 Not implemented Otoliths read only occasionally

Norway Barents Sea S. mentella and S. marinus


2 (3) Partially implemented

S. mentella and S. marinus read on a routine basis for stock assessment since 1990.

Norway Irminger Sea S. mentella Otoliths Break and burn

2 (3) Not implemented Otoliths read occasionally since 1993

Poland Irminger Sea S. mentella Otoliths Thin-section

1 Not implemented Reading since 2005

Russia Barents Sea S. mentella and S. marinus


1 Not implemented Reading since 1991. S. mentella read on a routine basis for stock assessment

Russia Irminger Sea S. mentella Scales 2 Not implemented Reading since 1980 Spain Flemish Cap S. marinus, S.

mentella and S. fasciatus

Otoliths Break and baked

1(2) Partially implemented

Reading since 1990. S. mentella read on a routine basis for stock assessment.

Canada Pacific stocks >30 rockfish species


9 Fully implemented Read on a routine basis for stock assessment

* Regular tests are performed in order to investigate the reliability of the reader.


Table 2 Summary of otoliths available and read 2006-2008.

Year Species Area 2006 2007 2008* Total % read

S. mentella Barents Sea/Norwegian Sea 3377 4522 3426 11325 23.1 Ice land and Greenland 1481 1258 62 2801 1.5 Irminger Sea complex 412 908 95 1415 9.2 Flemish Cap 4091 3112 4157 11360 9.2 Grand Banks of Newfoundland 3408 4593 1486 9487 0.0

S. marinus Barents Sea/Norwegian Sea 1652 3260 3307 8219 42.9 Ice land and Greenland 880 1083 1963 24.8 Irminger Sea complex 9290 9681 9451 28422 0.0 Flemish Cap 312 312 0.0

S. fasciatus Flemish Cap 1652 3260 3307 8219 0.0 Total redfish 25394 29220 22296 76910 14.2

Rockfish Canada West coast – British Columbia 26041 26555 6376 58972 7.8

* provisional

2.1 Otolith exchanges 2007-2008

During early 2007 to mid-2008, several sets of redfish otoliths were exchanged and read to compare age reading results for several stocks. Table 2.1 provides information on these ex-change sets. The results of the comparisons between readers and precision statistics are given in the following section. The methodology of the analysis of age readings follows standard procedures for precision and bias estimates and graphical presentations (Stransky et al. 2005a). A reading ‘confidence index’ (CI, see ICES 2006) was only applied by the Canadian and Spanish readers.

2.1.1 Demersal S. mentella from East Greenland

Exchange set 1 contained medium-sized S. mentella (>17 cm total length due to uncertain spe-cies determination below this size) from the continental shelf of East Greenland from a Ger-man research cruise in autumn 2000 (Table 2.1). One otolith of each pair in this set had been used in radiometric age validation (Stransky et al. 2005b). The Polish and Russian age read-ings (Table 2.1.1) were disregarded in the precision and bias estimation due to inconsistent results and likely non-conformity with the ageing criteria established in the last workshop (ICES 2006).

Age estimates varied between 2 and 11 years (Table 2.1.1). The percent agreement (PA) val-ues ranged between 0-75% (average 34.2%), the average coefficient of variation (CV) was 19.3% and the average percent error (APE) was 13.7% (Table 2.1.1).

The age-bias plots (Fig. 2.1.1) showed a slight non-linear bias for all countries, with a relative slight under-estimation of age by Canada and Iceland and a slight relative overestimation by Norway.


Table 2.1 Overview of the redfish otolith exchanges carried out 2007-2008.

Set Country providing samples

Species Area Type1 Gear2 Sampling period

Structure3 Method4 No. of otoliths

Canada Iceland Poland Norway Russia Spain Comments

1 Germany S. mentella ICES XIVb res dem 2000 oto bb 30 x x x x x x radiometrically validated

2 Germany S. mentella ICES XII ; NAFO 1F/2H/2J

res pel 2001 oto bb 30 x x x x x x radiometrically validated

3 Iceland S. marinus ICES Va com dem 2000 oto bb 30 x x x x - x radiometrically validated

4 Iceland S. marinus ICES Va res dem 2005 oto bb 30 x x x x - x

5 Russia S. mentella ICES XII ,XIV; NAFO 1F

com pel 1999-2005 oto bb 100 x x x x x x

6 Russia S. mentella ICES II (Norw.Sea) com pel 2006 oto/sc bb 30 x x x x x x

7 Spain S. marinus NAFO 3M res dem 2007 oto bb/bk 30 x x x x - x

8 Spain S. mentella NAFO 3M res dem 2007 oto bb/bk 30 x x x x - x

9 Spain S. fascia tus NAFO 3M res dem 2007 oto bb/bk 30 x x x x - x

Extra samples (origi nally not fore seen in ICES 2006):

E1 Norway S. mentella ICES II (Norw.Sea) com(re s) pel 2007 oto bb/ts 23 x x x x x -

E2 Spain Sebastes spp. (juveniles)

NAFO 3M res dem 2007 oto bb/bk 10 x x x x - x

1 Res= research, com= commercial 2 dem=demersal, pel= pelagic 3 oto= otoliths, sc=scales 4 bb = break-and-burn; bk = break-and-baked; ts = thin-section


Table 2.1.1 Age reading results for exchange set 1 (S. mentella East Greenland).

Fish information Age reading estimates (years) by nation1 and average Precision statistics

FishNo Length

(cm) Sex Canada Iceland Norway Poland Russia

1 Russia

2 Spain Avg. age PA CV APE

WH221-1034 19.5 M 3 3 5 (12) (3) (3) 5 4 0% 29% 25%

WH221-1050 18.5 M 2 2 5 (11) (8) (5) 4 3 0% 46% 42%

WH221-1028 20.5 F 4 4 5 (13) (6) (4) 6 5 25% 20% 15%

WH221-1044 17.5 M 5 4 5 (11) (8) (5) 5 5 75% 11% 5%

WH221-1060 21.5 M 4 5 6 (12) (5) (8) 4 5 25% 20% 15%

WH221-1077 23.5 F 6 5 7 (15) (10) (8) 4 6 25% 23% 17%

WH221-1055 27.5 M 5 6 8 (15) (9) (6) 6 6 50% 20% 13%

WH221-1058 26.5 M 7 6 8 (15) (8) (9) 6 7 25% 14% 11%

WH221-1092 25.5 M 4 6 8 (14) (9) (9) 6 6 50% 27% 17%

WH221-1001 28.5 M 6 8 8 (16) (10) (5) 8 8 75% 13% 6%

WH221-1008 22.5 M 7 7 8 (13) (8) (9) 9 8 25% 12% 9%

WH221-1024 24.5 F 8 8 9 (12) (8) (6) 6 8 50% 16% 9%

WH221-1003 29.5 F 6 9 9 (15) (10) (6) 10 9 50% 20% 11%

WH221-1065 31.5 M 7 8 10 (16) (13) (10) 8 8 50% 15% 9%

WH221-1072 32.5 F 7 9 10 (16) (10) (12) 8 9 25% 15% 11%

WH221-1064 30.5 M 8 10 12 (17) (12) (10) 7 9 0% 24% 19%

WH221-1146 33.5 F 8 9 11 (17) (11) (11) 9 9 50% 14% 8%

WH221-1178 34.5 F - 9 9 (17) (9) (11) 12 10 0% 17% 13%

WH221-1114 35.5 F 11 10 11 (17) (10) (10) 12 11 50% 7% 5%

Average 34.2% 19.3% 13.7%

1 Large ly deviating age estimates were not considered for precision estimates (indicated with parentheses), as non-conformity with age ing crite ria (ICES 2006) is very like ly.


Figure 2.1.1 Age-bias plots for exchange set 1 (S. mentella East Greenland) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given average age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

2.1.2 Pelagic S. mentella from the Irm inger Sea and adjacent waters

Exchange set 2 contained adult pelagic S. mentella (length range 26-45 cm) from the Irminger Sea, collected on the German part of the ‘International Hydroacoustic-Trawl Survey on Red-fish’ in June/July 2001 (Table 2.1). One otolith of each pair in this set had been used in radi-ometric age validation (Stransky et al. 2005b). The Russian age readings (Table 2.1.2) were disregarded in the precision and bias estimation due to inconsistent results and likely non-conformity with the ageing criteria established in the last workshop (ICES 2006).

Age estimates varied between 7 and 64 years (Table 2.1.2). The Canadian reader gave confi-dence indices (CI) of ‘fair’, ‘fairly poor’ or ‘poor’ in almost all cases (but 2 times ‘fairly good’). The CI distribution of the Spanish reader was similar, with a slightly higher overall confi-dence (5 times ‘fairly good’ as the best score). The PA values ranged between 0-60% (average 17.3%), the average CV was 17.5% and the APE was 12.9% (Table 2.1.2).

The age-bias plots (Fig. 2.1.2) showed high variation in age estimates for fish older than 20 years. The highest deviation from the overall reading results were observed for Poland, showing a severe relative under-estimation of fish that were estimated >30 years by most other readers, but <30 years by the Polish reader. The Canadian reader showed the highest positive deviation from the overall average of all readers for ages >25 years. The Spanish reader produced highly variable results for some of the age-groups. The Norwegian and Ice-landic readers, being most familiar with otoliths from this stock, had the lowest bias.

0

5

10

15

20

0 5 10 15 20

Mea

n ag

e +/

-2sd

Mean age all readers (years)

Canada

0

5

10

15

20

0 5 10 15 20

Mea

n ag

e +/

-2sd


Iceland

0

5

10

15

20

0 5 10 15 20

Mea

n ag

e +/

-2sd


Norway

0

5

10

15

20

0 5 10 15 20M

ean

age

+/-2

sd


Spain


Table 2.1.2 Age reading results for exchange set 2 (pelagic S. mentella Irminger Sea).

Fish information Age reading estimates (years) by nation1, confidence index (CI) and average age Precision statistics

FishNo Length

(cm) Sex Canada CI Iceland Norway Poland Russia 1 Russia 2 Spain CI Avg. age PA CV AP E

WH229-1002 44.5 F 64 FP 42 43 27 (17) (15) 23 FP 40 0% 41% 30%

WH229-1005 42.5 F 33 F 29 26 28 (14) (16) 20 FP 27 0% 18% 13%

WH229-1007 45.5 M 45 FP 41 33 31 (16) (17) 18 P 34 0% 31% 22%

WH229-1014 43.5 F 36 FP 27 27 26 (13) (16) 21 FP 27 40% 20% 12%

WH229-1025 41.5 M 39 F 34 32 27 (17) (14) 24 FP 31 0% 19% 15%

WH229-1079 30.5 F 9 F 9 10 11 (12) ( ) 10 F 10 40% 9% 6%

WH229-1104 36.5 M 29 FP 30 27 25 (15) (13) 20 F 26 0% 15% 12%

WH229-1109 37.5 M 26 F 28 23 23 (14) (11) 15 FP 23 40% 22% 14%

WH229-1119 31.5 M 26 F 25 20 18 (11) (13) 19 FP 22 0% 17% 15%

WH229-1122 36.5 M 38 F 39 28 25 (13) (14) 33 FP 33 20% 19% 15%

WH229-1128 33.5 M 29 F 29 22 25 (15) (11) 26 P 26 20% 11% 8%

WH229-1135 38.5 M 40 P 40 33 24 (18) (15) 39 P 35 0% 20% 15%

WH229-1138 29.5 F 10 FP 9 14 12 (11) (10) 13 P 12 20% 18% 13%

WH229-1139 34.5 M 34 FG 35 27 25 (18) (12) 32 F 31 0% 14% 12%

WH229-1140 37.5 F 24 F 26 20 26 (17) (14) 23 F 24 20% 10% 8%

WH229-1143 31.5 F 11 F 11 12 12 (11) (10) 11 FG 11 60% 5% 4%

WH229-1151 38.5 M 47 F 40 33 25 (15) (11) 35 FP 36 0% 23% 17%

WH229-1153 39.5 F 40 FP 37 26 25 (16) (15) 39 P 33 0% 22% 19%

WH229-1156 33.5 F 16 F 18 17 18 (14) (13) 16 F 17 20% 6% 5%

WH229-1160 27.5 M 8 F 8 8 12 (8) (9) 7 FG 9 0% 23% 18%

WH229-1173 39.5 F 50 FP 42 32 28 (19) (15) 38 P 38 20% 23% 17%

WH229-1177 35.5 M 37 F 34 27 26 (14) (14) 34 FP 32 0% 15% 13%

WH229-1190 32.5 M 12 FP 13 13 15 (15) (11) 10 FG 13 40% 14% 9%

WH229-1195 28.5 F 9 F 10 10 13 (10) (9) 9 FG 10 40% 16% 10%

WH229-1200 40.5 F 49 FP 51 41 27 (19) (15) 44 FP 42 0% 22% 16%

WH229-1260 32.5 F 11 F 11 12 13 (13) (10) 9 FG 11 40% 13% 9%

WH229-1265 34.5 M 25 F 25 21 19 (16) (15) 22 F 22 20% 12% 9%

WH229-1274 35.5 M 25 F 26 24 18 (14) (13) 23 F 23 20% 13% 10%

WH229-1275 26.5 M 8 FG 9 10 9 (7) (9) 7 F 9 40% 13% 9%

WH229-1392 40.5 M 48 FP 49 38 27 (16) (16) 40 P 40 20% 22% 16%

Average 17.3% 17.5% 12.9%



Figure 2.1.2 Age-bias plots for exchange set 2 (pelagic S. mentella Irminger Sea) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given average age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

2.1.3 S. mar inus from Iceland

Exchange set 3 consisted of S. marinus otoliths collected around Iceland on commercial trips in March 2000 (Table 2.1). The other-sided otolith of each pair in this set had been used in radiometric age validation (Stransky et al. 2005b).

Age estimates ranged from 1 to 44 years (Table 2.1.3). Regarding reading confidence, most of the individuals aged by the Canadian reader and Spanish reader 1 were given a CI of ‘fairly good’ and ‘fair’. None of the Canadian readings were given a CI of ‘good’, but the Spanish reader 1 assigned a ‘good’ to 6 individuals. The PA values ranged between 0-100% (average 34.6%), the average CV was 20.3% and the APE was 14.7% (Table 2.1.3).

Good correspondence between readers is visible in the age-bias plots (Fig. 2.1.3), with some moderate variation for ages >20 years.

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Mea

n ag

e +/

-2sd


Canada

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Mea

n ag

e +/

-2sd


Iceland

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Mea

n ag

e +/

-2sd


Norway

0

10

20

30

40

50

60

0 10 20 30 40 50 60

Mea

n ag

e +/

-2sd


Poland

0

10

20

30

40

50

60

0 10 20 30 40 50 60M

ean

age

+/-2

sdMean age all readers (years)

Spain


Table 2.1.3 Age reading results for exchange set 3 (S. marinus from Iceland).

Fish information Age reading estimates (years) by nation, confidence index (CI) and average age Precision statistics

FishNo Length

(cm) Sex Canada CI Iceland Norway Poland Spain

1 CI Spain

2 Avg. age PA CV AP E

TB_19_3 31 F 9 F 10 11 10 8 FG 10 10 50% 11% 7% TB_19_4 40 M 12 FG 12 13 12 11 F 10 12 50% 9% 6% TB_19_9 33 M 12 FG 13 12 12 12 F 11 12 67% 5% 3% TB_19_11 44 F 13 F 15 13 13 13 F 9 13 67% 16% 8% TB_19_12 29 M 9 F 9 9 10 10 F 7 9 50% 12% 7% TB_19_14 37 M 10 FP 12 10 11 9 F 9 10 33% 11% 8% TB_19_25 50 F 18 F 15 13 19 16 FP 14 16 17% 15% 11% TB_19_29 32 M 10 F 10 10 10 8 FG 9 10 67% 9% 5% TB_36_2 19 M 4 FG 4 4 7 4 G 8 5 0% 36% 30% TB_36_5 24 M 5 FG 5 5 7 4 FG 5 5 67% 19% 10% TB_36_7 14 U 3 F 2 2 4 1 G 2 40% 48% 40% TB_37_1 25 M 5 FG 6 5 5 5 F 5 5 83% 8% 3% TB_37_5 28 F 8 FG 8 7 8 7 G 8 8 67% 7% 4% TB_37_16 17 M 4 FG 4 4 4 3 G 3 4 67% 14% 8% TB_37_19 15 M 4 FG 4 4 4 3 FG 3 4 67% 14% 8% TB_46_7 18 F 4 F 3 5 3 2 FG 2 3 33% 37% 28% TB_46_8 26 F 6 FG 6 6 6 6 FG 6 6 100% 0% 0% TB_47_2 20 M 6 F 4 7 6 6 F 6 6 67% 17% 8% TB_47_6 11 U 4 F 3 3 3 2 FG 2 3 50% 27% 17% TB_47_7 21 F 6 FP 5 6 12 5 G 5 7 0% 42% 31% TB_48_2 12 U 4 F 2 3 4 1 F 1 3 17% 55% 39% TB_48_6 36 F 16 FG 14 15 16 11 F 9 14 17% 21% 15% TB_48_8 35 M 12 F 12 13 12 11 F 11 12 50% 6% 4% TB_49_3 22 M 6 FG 6 8 8 6 F 5 7 0% 19% 17% TB_49_5 16 F 5 FP 3 5 6 3 FG 7 5 33% 33% 23% TB_49_6 27 F 10 FP 7 11 9 8 F 7 9 17% 19% 15% TB_49_8 39 F 12 FG 12 13 12 10 FG 9 11 0% 13% 12% TB_49_9 13 U 4 F 3 3 4 2 FG 1 3 33% 41% 28% TB_49_14 30 F 10 FG 10 13 10 9 F 7 10 50% 20% 12% TB_51_1 43 F 19 F 15 14 14 16 FG 14 15 17% 13% 9% TB_51_2 38 M 12 F 11 12 12 12 F 10 12 67% 7% 4% TB_51_3 49 M 24 FP 20 18 26 16 F 14 20 17% 24% 18% TB_52_1 55 M 29 FP 23 16 27 22 FP 25 24 0% 19% 14% TJ_47_8 23 F 7 F 6 5 7 5 F 28 10 0% 93% 63% TJ_55_5 34 U 11 FG 11 10 13 10 G 13 11 33% 12% 9% TJ_65_7 41 F 16 FP 12 14 15 14 25% 12% 9% TJ_66_1 53 M 44 FP 39 30 29 37 P 11 32 0% 37% 26% TJ_103_1 48 M 32 F 30 27 27 28 FP 29 0% 8% 22% TJ_107_1 42 F 14 F 13 13 13 13 FG 14 13 67% 4% 3% TJ_119_1 45 F 16 FP 16 18 17 14 F 22 17 17% 16% 11% TJ_120_1 47 F 17 FP 16 18 16 5 14 0% 37% 29% TJ_120_4 46 F 16 F 15 16 14 14 F 12 15 17% 10% 8% TJ_120_7 51 F 17 FG 16 15 16 14 FP 14 15 17% 8% 7% TJ_128_2 52 M 26 F 23 28 25 24 P 36 27 0% 18% 12% TL_125_1 54 F 17 20 18 23 20 25% 14% 10%

Average 34.6% 20.3% 14.7%


Figure 2.1.3 Age-bias plots for exchange set 3 (S. marinus from Iceland) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given average age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

2.1.4 S. mar inus from Iceland, 1990 yearc lass

The otoliths in exchange set 4 were selected from S. marinus samples taken around Iceland in 2005 and were identified as belonging to the strong 1990 year-class by the Icelandic reader (being most experienced with this stock). These fish had a total length of 31-44 cm.

Consequently, the Icelandic reader assigned age 15 to all individuals. The other readers’ age estimates varied between 11 and 21 years (Table 2.1.4). The Spanish reader had markedly less confidence in the readings (CI ‘fairly poor’ or ‘poor’ in half of the cases) than the Canadian reader (‘fairly poor’ in only 4 cases, representing the least confidence). There was only one fish for which all readers agreed at that age, thus reaching 100% agreement. For the other fish, the PA ranged from 0-80%, and the overall average was 33.3%. The Canadian reader assigned age 15 to two thirds of the individuals and mostly ages >15 years to the other fish, while the other readers only assigned 16-31% of individuals to age 15. The two Spanish read-ers mostly assigned ages <15 years to the remaining individuals, whereas the Polish reader had a tendency to assign ages >15 years (Fig. 2.1.4). 80% of the Norwegian readings were in the range 15 years ±1 year.

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Canada

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Iceland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Norway

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Poland

0

10

20

30

40

50

0 10 20 30 40 50M

ean

age

+/-2

sd


Spain_R1

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Spain_R2


Table 2.1.4 Age reading results for exchange set 4 (S. marinus from Iceland, 1990 year-class)

Fish information Age reading estimates (years) by nation, confidence index (CI) and

average age Precision statistics

FishNo Length

(cm) Sex Canada CI Iceland Norway Poland Spain

1 CI Spain

2 Avg. age PA

1(2) 36 M 15 F 15 16 15 14 F 13 15 50%

2(3) 38 M 18 FP 15 19 18 13 FG 16 17 0%

3(2) 37 F 17 F 15 18 18 18 FP 15 17 17%

4(3) 36 M 15 F 15 14 16 15 FG 13 15 50%

5(6) 34 M 16 FG 15 16 16 12 P 16 15 17%

6(1) 36 F 15 FG 15 16 16 11 FP 15 40%

7(3) 37 F 15 F 15 15 15 14 F 13 15 67%

8(4) 35 F 15 FG 15 15 15 15 F 15 15 100%

9(2) 34 M 15 FG 15 15 16 15 F 15 80%

10(8) 37 F 15 F 15 13 15 12 FP 13 14 0%

11(11) 31 M 15 F 15 14 17 13 FP 17 15 33%

12(3) 35 M 15 F 15 16 17 14 F 16 16 33%

13(4) 36 M 18 F 15 15 17 15 FP 15 16 0%

14(15) 37 F 15 FG 15 16 14 14 FP 15 15 50%

15(2) 41 F 15 FG 15 16 15 15 F 14 15 67%

16(5) 40 F 15 F 15 15 15 11 F 11 14 0%

17(8) 39 F 17 F 15 16 16 15 FP 11 15 33%

18(10) 40 M 15 F 15 16 16 14 F 14 15 33%

19(2) 35 M 15 F 15 14 17 15 F 13 15 50%

20(3) 39 M 14 F 15 15 19 13 P 17 16 0%

21(5) 40 F 16 FP 15 16 18 16 P 13 16 50%

22(9) 44 M 15 F 15 15 16 16 P 15 60%

23(11) 34 M 15 F 15 14 16 15 F 12 15 50%

24(1) 37 F 21 F 15 16 20 18

25(2) 40 F 15 FP 15 16 18 17 FP 14 16 17%

26(10) 36 M 14 F 15 16 16 15 FP 13 15 33%

27(11) 37 F 14 F 15 13 17 13 FG 12 14 17%

28(17) 41 M 14 FP 15 16 17 14 P 13 15 17%

29(20) 40 F 15 F 15 17 17 16 F 14 16 17%

30(21) 40 F 15 FG 15 18 14 16 FP 16 20%

Average 33.3%


Figure 2.1.4 Age readings for exchange set 4 (S. marinus from Iceland, 1990 year-class) by nation.

2.1.5 Pelagic S. mentella from the Irm inger Sea and adjacent waters , Russ ian sam-ples

Exchange set 5 contained adult pelagic S. mentella (length range 25-41 cm) from the Irminger Sea, collected by Russia in 1999-2005 (Table 2.1). The Russian age readings (Table 2.1.5) were disregarded in the precision and bias estimation due to inconsistent results and likely non-conformity with the ageing criteria established in the last workshop (ICES 2006).

Age estimates varied between 7 and 60 years (Table 2.1.5). The Canadian reader assigned a CI of ‘fairly poor’ to more than half of the cases, ‘poor’ to 10%, and only 6 out of 103 otoliths were read with ‘fairly good’ as the best confidence. The PA values ranged between 0-80% (average 17.6%), the average CV was 20.6% and the APE was 15.0% (Table 2.1.5).

The age-bias plots (Fig. 2.1.5) showed high variation in age estimates for all readers. A consi-derable relative age overestimation of fish >20 years was observed for the Canadian reader, while the Polish reader showed the opposite trend and estimated <30 years for those fish that were estimated 30-40 years by most other readers (Fig. 2.1.5). Consistent slight relative unde-restimation was visible in the results of the Icelandic reader. The Norwegian and Spanish readers had the lowest overall bias.

10

11

12

13

14

15

16

17

18

19

20

21

22

0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32

Age

(yea

rs)

Otolith no.

CAN

ICE

NOR

POL

SPA_R1

SPA_R2


Table 2.1.5 Age reading results for exchange set 5 (pelagic S. mentella from the Irminger Sea and adjacent waters, Russian samples)

Fish information Age reading estimates (years) by nation1, confidence index (CI)

and average age Precision statistics

FishNo Length

(cm) Sex Canada CI Iceland Norway Poland Russia Spain

2 Avg. age PA CV APE

M0012-17_10 36 F 27 F 22 26 19 (16) 20 23 0% 16% 13%

M0012-17_17 28 F 10 F 9 9 10 (9) 14 10 40% 20% 12%

M0012-23_13 35 F 32 FG 24 26 24 (14) 24 26 20% 13% 9%

M0012-31_6 35 F 15 FP 17 14 18 (14) 13 15 20% 13% 11%

M0012-31_28 27 M 8 FG 9 9 10 (8) 9 9 60% 8% 4%

M0012-31_30 25 M 6 FG 6 8 8 (6) 8 7 0% 15% 14%

M0012-44_18 27 M 7 FG 7 8 9 (8) 7 8 20% 12% 10%

M0058-37_3 34 M 20 F 17 14 21 (12) 20 18 0% 16% 13%

M0058-37_8 39 F 40 FP 25 28 25 (16) 30 30 20% 21% 15%

M0058-37_15 40 F 41 FP 18 20 22 (16) 29 26 0% 36% 28%

M0058-37_26 39 F 27 FP 16 24 24 (15) 22 23 0% 18% 12%

M0058-37_34 39 F 17 FP 17 26 21 (17) 23 21 20% 19% 14%

M0058-37_40 31 M 13 FP 13 12 12 (12) 13 13 60% 4% 3%

M0058-37_44 30 M 13 FG 12 11 10 (11) 9 11 20% 14% 11%

M0058-37_45 34 M 35 P 23 24 17 (13) 24 25 0% 26% 18%

M0058-37_49 35 M 9 FG 10 13 12 (14) 9 11 0% 17% 15%

M0058-38_1 32 F 13 FP 11 17 10 (13) 15 13 20% 22% 17%

M0058-38_4 34 F 10 FP 13 13 12 (12) 10 12 20% 13% 10%

M0058-38_5 39 F 30 FP 18 24 16 (17) 17 21 0% 28% 23%

M0058-38_6 32 M 13 FP 12 11 10 (12) 11 11 40% 10% 7%

M0058-38_7 34 F 13 FP 13 13 14 (14) 11 13 60% 9% 5%

M0058-38_9 36 F 31 F 25 25 22 (16) 28 26 0% 13% 10%

M0058-38_11 30 M 10 FP 11 10 10 (12) 12 11 20% 8% 7%

M0058-38_13 29 F 10 FP 10 10 9 (10) 10 75% 5% 3%

M0058-38_16 34 F 14 F 14 12 13 (13) 12 13 20% 8% 6%

M0058-38_19 38 F 30 P 23 18 21 (16) 22 23 20% 19% 13%

M0058-38_21 35 F 8 FP 12 13 16 (15) 9 12 20% 28% 20%

M0058-38_22 34 F 13 FP 14 15 18 (14) 15 15 40% 12% 8%

M0058-38_25 33 F 14 FP 13 13 19 (13) 14 15 0% 17% 13%

M0058-38_28 38 M 38 FP 30 26 27 (16) 33 31 0% 16% 12%

M0058-38_30 28 M 10 F 10 10 8 (10) 10 10 80% 9% 4%

M0058-38_31 39 F 40 FP 23 27 24 (17) 24 28 0% 26% 19%

M0058-38_32 32 F 8 FP 12 10 12 (12) 10 10 40% 16% 12%

M0058-38_38 33 F 9 F 10 11 12 (12) 11 11 40% 11% 7%

M0058-38_44 35 M 37 F 22 24 21 (15) 13 23 0% 37% 24%

M0058-46_1 37 M 35 P 21 19 27 (17) 14 23 0% 35% 27%

M0058-46_4 35 M 40 F 25 23 25 (15) 16 26 0% 34% 22%

M0058-46_6 29 M 7 FP 9 11 10 (10) 10 9 20% 16% 13%

M0058-46_7 30 M 7 FP 10 11 9 (9) 8 9 20% 18% 13%

M0058-46_11 37 F 32 FP 23 22 9 (10) 16 20 0% 42% 32%

M0058-46_12 29 M 6 FP 9 8 24 (15) 7 11 0% 69% 49%


M0058-46_16 36 M 32 F 24 24 10 (12) 24 23 0% 35% 22%

M0058-46_21 31 M 7 FP 10 10 23 (17) 13 0% 57% 42%

M0058-46_23 37 F 30 F 22 25 24 (17) 23 25 20% 13% 9%

M0058-46_29 40 F 40 FP 19 23 24 (15) 17 25 0% 37% 26%

M0058-49_4 36 M 30 FP 15 19 10 (11) 19 19 40% 40% 25%

M0058-49_5 30 F 6 FP 10 10 10 (12) 8 9 0% 20% 16%

M0058-49_7 31 M 6 FP 10 11 21 (16) 14 12 0% 45% 33%

M0058-49_12 30 M 10 F 9 11 12 (11) 10 10 40% 11% 8%

M0058-49_17 31 F 13 FP 11 11 10 (12) 10 11 40% 11% 7%

M0058-49_21 33 M 11 FP 9 12 13 (12) 14 12 20% 16% 12%

M0058-49_23 39 F 14 FP 17 17 19 (17) 17 17 60% 11% 6%

M0058-49_28 33 M 31 FP 21 14 19 (14) 24 22 0% 29% 21%

M0058-49_35 31 M 10 FP 9 12 9 (11) 14 11 0% 20% 16%

M0058-49_38 37 F 35 P 20 23 23 (16) 20 24 0% 26% 18%

M0072-1_5 36 M 30 P 16 17 20 (15) 19 20 20% 27% 18%

M0072-1_11 38 M 55 FP 27 29 27 (17) 32 34 0% 35% 25%

M0072-1_14 38 M 35 FP 15 29 23 (16) 23 25 0% 30% 22%

M0072-1_21 37 M 35 FP 18 27 24 (16) 20 25 0% 27% 20%

M0072-9_11 38 M 38 P 30 28 24 (16) 22 28 20% 22% 16%

M0072-9_19 38 M 37 P 20 23 22 (16) 16 24 0% 34% 23%

M0072-9_30 28 F 8 F 8 9 9 (10) 10 9 40% 10% 7%

M0072-17_24 36 M 35 F 19 16 25 (15) 23 24 0% 31% 22%

M0072-24_4 33 M 14 F 12 13 16 (14) 14 14 40% 11% 7%

M0072-24_6 38 F 37 FP 21 28 26 (17) 31 29 0% 21% 15%

M0072-24_10 38 M 38 F 17 17 24 (16) 22 24 20% 37% 25%

M0072-24_29 41 F 36 FP 24 20 26 (17) 28 27 0% 22% 16%

M0072-24_31 32 M 11 FP 12 11 14 (12) 18 13 0% 22% 17%

M0072-24_35 28 M 7 F 8 8 9 (9) 8 50% 10% 6%

M0072-24_37 37 M 33 F 23 18 26 (15) 20 24 0% 24% 18%

M0072-24_45 40 F 32 F 21 22 27 (17) 17 24 0% 24% 19%

M0072-40_5 29 M 8 F 8 9 9 (9) 13 9 40% 22% 13%

M0072-40_9 32 F 14 F 10 12 10 (12) 12 12 40% 14% 10%

M0072-40_11 30 F 10 FP 9 10 10 (10) 11 10 60% 7% 4%

M0072-40_14 40 F 29 F 15 19 26 (16) 23 22 0% 25% 20%

M0072-40_21 32 M 14 F 11 13 15 (12) 14 13 20% 11% 9%

M0072-40_25 31 F 12 F 10 11 14 (11) 10 11 20% 15% 11%

M0072-40_34 29 M 10 FP 9 9 10 (10) 12 10 40% 12% 8%

M0072-40_35 32 M 12 FP 12 13 13 (11) 12 12 60% 4% 3%

M0059-5_2 37 M 37 FP 27 27 22 (16) 27 28 0% 20% 13%

M0059-5_3 37 F 26 FP 18 19 24 (13) 19 21 0% 17% 14%

M0059-5_6 35 M 28 F 22 24 23 (13) 18 23 20% 16% 10%

M0059-5_7 33 F 9 FP 10 13 13 (12) 11 0% 18% 16%

M0059-5_12 34 M 22 P 13 15 19 (13) 9 16 0% 33% 25%

M0059-5_15 35 F 13 F 13 12 18 (13) 15 14 0% 17% 13%

M0059-5_17 33 M 20 F 15 13 18 (11) 19 17 0% 17% 14%

M0059-5_23 34 F 9 FP 11 12 14 (11) 13 12 20% 16% 12%


M0059-19_32 36 F 17 F 16 16 22 (16) 17 18 0% 14% 11%

M0059-19_36 31 F 10 FP 11 12 (18) 11 50% 7%

M0059-19_42 37 M 30 P 21 22 25 (15) 17 23 0% 21% 16%

M0059-19_47 32 F 8 FP 12 12 18 (11) 12 12 60% 29% 17%

M0059-19_48 31 F 14 FP 11 12 12 (12) 14 13 0% 11% 9%

M0059-23_10 28 F 8 FP 9 9 9 (9) 9 9 80% 5% 2%

M0059-23_50 29 F 9 F 10 9 12 (10) 11 10 20% 13% 10%

M0059-35_14 33 F 11 FP 12 13 13 (12) 12 12 40% 7% 5%

M0059-35_27 29 F 7 F 10 10 11 (10) 20 12 0% 42% 30%

M0056-4_5 30 F 7 F 10 8 10 (10) 10 9 0% 16% 13%

M0056-4_6 36 F 25 FP 20 27 26 (15) 26 25 20% 11% 7%

M0056-4_11 36 M 35 FP 17 24 26 (15) 24 25 0% 26% 17%

M0056-4_27 38 F 33 F 25 28 26 (17) 31 29 0% 12% 10%

M0056-8_17 38 F 36 P 19 27 21 (17) 19 24 0% 30% 23%

M0056-58_17 39 M 60 FP 32 38 30 (17) 42 40 0% 30% 21%

M0056-58_23 38 M 53 P 23 28 27 (15) 35 33 0% 36% 26%

Average 17.6% 20.6% 15.0%



Figure 2.1.5 Age-bias plots for exchange set 5 (pelagic S. mentella from the Irminger Sea and adjacent wa-ters, Russian samples) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given a verage age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

2.1.6 Pelagic S. mentella from the Norwegian Sea, Russ ian samples

Exchange set 6 consisted of otoliths from adult pelagic S. mentella (length range 34-42 cm) from the Norwegian Sea, collected by Russia in 2006 (Table 2.1). The Russian age readings (Table 2.1.6) were disregarded in the precision and bias estimation due to inconsistent results and likely non-conformity with the ageing criteria established in the last workshop (ICES 2006).

Age estimates varied between 12 and 46 years (Table 2.1.6). Both the Canadian and the Span-ish reader had only ‘fair’ to ‘poor’ confidence in their readings, with 80% of the material as-signed ‘fairly poor’ or ‘poor’. The Spanish reader had ‘poor’ confidence in half of the cases. The PA values ranged from 0 to 40% (average 8.7%), the average CV was 21.7% and the APE was 15.7% (Table 2.1.6).

High positive bias was observed in the age-bias plots (Fig. 2.1.6) for the Canadian reader, assigning most individuals ages of >30 years, and moderate non-linear bias for the other readers.

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Canada

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Iceland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Norway

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Poland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Spain_R2


Table 2.1.6 Age reading results for exchange set 6 (pelagic S. mentella from the Norwegian Sea, Russian samples)

Fish information Age reading estimates (years) by nation1, confidence index (CI) and average age

Precision statistics

Fish No

Length (cm) Sex Canada CI Iceland Norway Poland Russia

Spain 1 CI

Spain 2

Avg. age PA CV APE

1 38.9 F 39 P 20 17 25 (17) 16 P 16 22 0% 40% 30%

2 38.4 F 36 FP 23 19 21 (17) 22 P 22 24 0% 26% 17%

3 37.9 F 37 FP 26 24 22 (16) 24 P 19 25 0% 24% 16%

4 36.9 F 46 FP 18 17 19 (15) 12 P 16 21 0% 58% 38%

5 38.9 F 39 FP 30 27 24 (17) 29 FP 27 29 17% 18% 11%

6 35.6 M 15 FP 16 16 20 (14) 15 16 40% 13% 8%

7 36.2 F 27 F 26 18 21 (14) 25 FG 21 23 0% 15% 13%

8 36.9 M 42 P 29 24 25 (15) 29 FP 24 29 33% 24% 16%

9 36.5 M 40 FP 22 25 21 (16) 24 26 0% 29% 33%

10 37.4 M 42 FP 27 28 24 (17) 28 P 30 30 17% 21% 14%

11 37.8 M 35 F 26 28 21 (16) 31 P 27 28 17% 17% 12%

12 40.1 F 34 FP 24 29 26 (21) 25 P 26 27 0% 13% 10%

13 39.9 F 35 FP 25 25 29 (20) 27 P 17 26 0% 22% 15%

14 34.3 M 22 F 15 15 19 (14) 15 FP 15 17 0% 18% 15%

15 37.0 M 44 FP 26 23 23 (16) 29 0% 35% 26%

16 38.5 M 42 P 25 29 25 (17) 23 FP 20 27 0% 28% 20%

17 37.5 M 36 FP 23 24 26 (17) 27 P 27 20% 19% 13%

18 37.5 M 36 FP 32 26 25 (18) 33 P 27 30 0% 15% 13%

19 35.1 M 38 FP 27 25 19 (15) 30 FP 29 28 0% 22% 15%

20 35.5 M 34 FP 24 25 23 (16) 28 F 26 27 0% 15% 11%

21 37.7 F 27 F 21 23 25 (16) 20 P 18 22 0% 15% 12%

22 36.0 M 31 F 25 26 23 (15) 27 P 28 27 17% 10% 7%

23 37.5 F 38 FP 18 26 21 (16) 28 P 25 26 17% 27% 18%

24 36.5 M 19 F 15 16 22 (15) 15 F 15 17 0% 17% 14%

25 38.9 M 41 FP 27 28 27 (19) 30 FP 24 30 17% 20% 14%

26 42.0 F 38 FP 20 30 27 (20) 28 P 27 28 17% 21% 13%

27 38.5 M 39 FP 31 36 24 (17) 28 P 29 31 17% 18% 13%

28 38.7 F 36 FP 28 29 23 (19) 28 F 30 29 17% 14% 9%

29 37.4 M 33 FP 26 28 24 (18) 29 FP 27 28 17% 11% 8%

30 34.8 M 35 FP 25 20 (14) 22 FP 20 24 0% 26% 18%

Average 8.7% 21.7% 15.7%



Figure 2.1.6 Age-bias plots for exchange set 6 (pelagic S. mentella from the Norwegian Sea, Russian sam-ples) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given average age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

2.1.7 S. mar inus from Flem ish Cap

Exchange set 7 contained S. marinus otoliths from the Flemish Cap of 18-40 cm length, col-lected by Spain on a research survey in 2007 (Table 2.1).

Age estimates ranged from 4 to 21 years (Table 2.1.7). In 80% of the cases, the Canadian read-er had ‘fair’ or ‘fairly poor’ confidence in the readings, whereas the Spanish reader 1 (most familiar with this stock) had ‘good’ or ‘fairly good’ confidence in 60% of the cases. The PA values ranged from 0 to 67% (average 11.7%), the average CV was 25.6% and the APE was 20.1% (Table 2.1.7).

Figure 2.1.7 shows moderate non-linear bias for all readers, with the Polish reader and Span-ish reader 2 having the highest variation in age estimates. Disregarding the Polish results (and re-shaping the plots between 10-15 years), the bias in the Icelandic, Norwegian and Spanish (R1) readers’ is relatively minor.

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Canada

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Iceland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Norway

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Poland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Spain_R1

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Spain_R2


Table 2.1.7 Age reading results for exchange set 7 (S. marinus from Flemish Cap)

Fish information Age reading estimates (years) by nation, confidence index (CI) and aver-

age age Precision statistics Fish No

Length (cm) Sex Canada CI Iceland Norway Poland

Spain 1 CI

Spain 2

Avg. age PA CV APE

17_3 34 M 12 FP 11 18 14 10 FP 13 13 17% 22% 15%

17_4 37 M 18 F 15 21 18 16 FG 16 17 0% 12% 10%

17_7 38 F 17 F 16 17 18 15 FG 13 16 17% 11% 8%

17_10 33 M 18 F 14 16 15 14 F 10 15 17% 18% 12%

17_11 39 F 15 FP 15 16 16 16 F 16 16 67% 3% 2%

17_12 33 M 15 F 14 15 16 12 P 15 15 50% 10% 6%

52_2 28 M 8 FP 7 13 14 7 F 6 9 0% 37% 31%

52_3 37 M 16 FP 17 14 17 15 P 8 15 17% 23% 14%

52_4 30 M 9 F 7 13 13 8 FG 9 10 0% 26% 22%

52_7 31 M 12 F 7 13 12 9 FG 9 10 0% 23% 20%

52_9 27 M 14 FP 6 11 14 7 FG 8 10 0% 35% 30%

52_18 22 M 5 F 5 10 6 5 G 16 8 0% 57% 44%

52_26 32 F 8 FG 7 12 11 7 G 8 9 0% 24% 20%

52_27 31 F 9 FG 8 9

52_28 32 F 7 FG 7 7

52_31 25 F 8 FP 8 11 12 7 G 8 9 0% 22% 19%

52_33 29 F 8 FG 6 7 0%

52_34 24 F 8 F 7 11 11 7 FG 6 8 17% 26% 21%

52_35 26 F 8 F 8 9 9 8 G 9 9 50% 6% 6%

52_37 20 F 5 F 4 8 9 5 G 6 6 17% 31% 25%

52_38 23 F 5 G 5

52_40 19 F 5 F 5 5 7 6 G 5 6 17% 15% 14%

52_43 21 F 5 F 6 7 11 9 8 0% 32% 25%

52_44 18 F 4 F 5 6 8 5 G 8 6 17% 28% 22%

71_1 40 M 16 F 15 17 16 18 P 20 17 17% 11% 8%

71_7 34 F 12 F 14 17 18 13 F 14 15 0% 16% 13%

71_9 36 F 9 FG 9 15 21 10 FG 11 13 0% 38% 29%

83_1 36 F 7 F 8 13 20 12 0% 50% 38%

83_2 35 F 7 F 8 13 18 12 0% 44% 33%

84_3 35 F 8 FP 9 15 21 9 F 8 12 0% 45% 36%

Average 11.7% 25.6% 20.1%


Figure 2.1.7 Age-bias plots for exchange set 7 (S. marinus from Flemish Cap) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given average age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Canada

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Iceland

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Norway

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Poland

0

5

10

15

20

25

0 5 10 15 20 25M

ean

age

+/-2


Spain_R1

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Spain_R2


2.1.8 S. mentella from Flem ish Cap

Exchange set 8 contained S. mentella otoliths from the Flemish Cap of 18-40 cm length, col-lected by Spain on a research survey in 2007 (Table 2.1).


Fig. 2.1.8 shows only slight bias for most of the readers, with the Polish reader having posi-tive and the Spanish reader 2 having negative bias. The Canadian results had the lowest var-iation and bias.

Table 2.1.8 Age reading results for exchange set 8 (S. mentella from Flemish Cap)

Fish information Age reading estimates (years) by nation, confidence index (CI) and

average age Precision statistics Fish No


Spain 1 CI

Spain 2

Avg. age PA CV APE

9_4 18 M 4 FP 3 6 9 5 F 5 20% 43% 32%

19_6 37 F 14 FG 15 14 15 13 F 12 14 33% 8% 6%

19_7 38 F 16 F 16 15 18 14 F 13 15 17% 11% 9%

44_1 28 F 8 F 7 7 14 7 G 6 8 17% 36% 23%

44_2 32 F 12 F 12 14 13 12 P 6 12 50% 24% 13%

44_3 33 F 12 FG 13 12 13 11 FG 9 12 33% 13% 8%

44_4 30 F 9 F 8 11 11 7 G 7 9 17% 21% 17%

44_5 34 F 18 F 18 17 18 15 G 15 17 17% 9% 7%

44_6 30 F 9 F 9 10 12 7 FG 7 9 33% 21% 15%

44_7 33 F 13 F 12 12 13 10 G 9 12 33% 14% 10%

44_8 36 F 11 FG 12 12 14 10 F 12 40% 13% 8%

44_9 31 F 10 F 9 10 11 9 FG 6 9 33% 19% 13%

44_13 31 F 9 F 9 11 11 8 FG 7 9 33% 17% 13%

44_14 32 F 10 FG 8 9 0%

44_16 28 F 8 FP 7 11 10 9 FG 7 9 17% 19% 15%

44_19 26 F 7 FP 8 10 12 8 F 6 9 0% 26% 20%

44_20 21 F 6 FP 5 7 7 7 FG 6 6 33% 13% 11%

44_22 26 F 7 F 7 7 7 8 G 6 7 67% 9% 5%

44_23 22 F 7 FG 6 8 8 7 G 7 7 50% 11% 7%

44_24 27 M 9 F 9 8 11 7 FG 7 9 33% 18% 13%

44_30 27 M 8 F 8 9 10 7 G 6 8 33% 18% 13%

44_32 23 M 11 FG 7 9 14 8 G 7 9 17% 29% 22%

44_33 29 M 8 FG 7 9 11 8 FG 8 9 17% 16% 13%

44_34 25 M 7 F 6 8 12 7 FG 6 8 17% 29% 21%

44_35 24 M 8 F 7 9 12 8 FG 8 9 17% 20% 15%

44_36 29 M 10 F 8 10 13 9 FG 8 10 33% 19% 13%

44_42 19 M 7 FP 4 9 12 6 FG 6 7 17% 38% 29%

44_43 20 M 6 F 6 6 100%

75_19 35 F 17 FG 16 16 20 16 FG 16 17 17% 10% 7%

75_22 34 F 13 F 12 12 16 13 FG 11 13 33% 13% 9%

Average 29.2% 19.2% 13.8%


Figure 2.1.8 Age-bias plots for exchange set 8 (S. mentella from Flemish Cap) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given average age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Canada

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Iceland

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Norway

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Poland

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Spain_R1

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Spain_R2


2.1.9 S. fasc iatus from Flemish Cap

Exchange set 9 contained S. fasciatus otoliths from the Flemish Cap of 18-36 cm length, col-lected by Spain on a research survey in 2007 (Table 2.1).


Fig. 2.1.9 shows only slight bias for the Canadian, Icelandic and Norwegian readers, and moderate bias for the other readers. The Polish readings had a moderate positive bias and the Spanish (both readers) a negative.

Table 2.1.9 Age reading results for exchange set 9 (S. fasciatus from Flemish Cap)

Fish information Age reading estimates (year s) by nation, confidence index (CI) and average age

Precision statistics Fish No

Length (cm)

Sex Canada CI Iceland Norway Poland Spain 1

CI Spain 2

Avg. age

PA CV APE 21_1 28 M 11 FP 9 11 14 9 FG 9 11 33% 19% 14%

21_2 29 M 15 F 15 14 19 11 FG 11 14 17% 21% 15%

21_5 33 F 12 F 12 14 17 13 FP 11 13 17% 16% 12%

21_6 32 F 12 FP 11 12 14 10 F 10 12 33% 13% 10%

21_8 28 F 9 F 11 12 14 9 FG 9 11 17% 19% 15%

21_10 34 F 12 F 13 13 16 11 G 11 13 33% 15% 10%

21_12 31 F 11 FP 10 11 14 13 FG 13 12 0% 13% 11%

21_13 35 F 12 F 12 12 14 13 F 12 13 17% 7% 6%

44_3 24 M 14 FP 9 11 11 7 F 9 10 0% 24% 18%

44_4 21 M 8 F 7 8 9 6 G 6 7 17% 17% 14%

44_12 19 M 6 F 6 8 10 6 FG 6 7 0% 24% 19%

44_14 22 M 9 FP 7 10 11 6 FG 5 8 0% 30% 25%

44_15 23 M 10 FP 8 10 13 6 F 6 9 0% 31% 24%

44_17 25 M 13 FP 8 10 16 7 FG 7 10 17% 36% 28%

44_19 26 M 12 FP 11 12 13 8 FG 7 11 17% 23% 17%

44_20 26 M 7 G 8 8 50%

44_21 27 M 14 F 12 13 15 8 G 8 12 17% 26% 19%

44_22 27 M 13 F 12 12 13 12 F 9 12 50% 12% 7%

44_23 30 F 13 F 12 13 14 9 G 9 12 17% 19% 14%

44_25 30 F 10 F 12 12 15 10 F 10 12 33% 17% 13%

44_26 31 F 12 FG 13 14 16 10 FG 11 13 17% 17% 13%

44_27 34 F 17 FG 17 16 19 16 FP 11 16 33% 17% 10%

44_28 32 F 9 F 11 10 19 9 FG 10 11 17% 34% 21%

44_29 29 F 8 F 9 11 14

44_31 25 F 10 FP 9 11 14 8 FG 8 10 17% 23% 17%

44_39 20 F 5 G 5

44_40 18 F 7 FP 7 8 11 5 G 5 7 33% 31% 21%

60_3 29 F 9 FP 10 9 14 8 F 9 10 17% 22% 15%

60_5 35 F 19 F 17 13 18 17 F 13 16 0% 16% 14%

72_2 36 F 13 F 13 13 15 13 F 14 14 17% 6% 6%

111_1 33 F 17 FG 16 16 18 16 FG 12 16 50% 13% 7%

Average 20.1% 19.9% 14.9%


Figure 2.1.9 Age-bias plots for exchange set 9 (S. fasciatus from Flemish Cap) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given average age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Canada

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Iceland

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Norway

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Poland

0

5

10

15

20

25

0 5 10 15 20 25M

ean

age

+/-2


Spain_R1

0

5

10

15

20

25

0 5 10 15 20 25

Mea

n ag

e +/

-2sd


Spain_R2


2.1.10 Pelagic S. mentella from the Norwegian Sea

Exchange set E1 contained otoliths from adult pelagic S. mentella (length range 27-43 cm) from the Norwegian Sea, collected by Norway in 2007 (Table 2.1). These were read by break-and-burn and thin-sections for methodological comparisons. The Russian age readings (Tables 2.1.10a and 2.1.10b) were disregarded in the precision and bias estimation due to in-consistent results and likely non-conformity with the ageing criteria established in the last workshop (ICES 2006).

For the broken-and-burnt otoliths, age estimates varied between 8 and 41 years (Table 2.1.10a), for the thin-sections between 10 and 42 years (Table 2.1.10b). The CI of the Canadian reader was almost exclusively ‘fair’ or ‘fairly poor’ for the broken-and-burnt otoliths and mostly ‘poor’ for the thin-sections. The PA values ranged from 0 to 50% (average 14.1%) for the broken-and burnt otoliths and from 0 to 75% (average 20.7%) for the thin-sections. The average CV was 16.0% (break-and-burn) and 15.0% (thin-sections), respectively, and the APE was 11.8% and 10.9%, respectively (Tables 2.1.10a and 2.1.10b).

The bias patterns were similar by country for broken-and-burnt (Figure 2.1.10a) and thin-sectioned (Fig. 2.1.10b) otoliths. The Icelandic and Norwegian readings had relatively low non-linear bias, while the Canadian readings had positive bias for ages >20 years and the Polish readings had negative bias especially between 25 and 35 years (overall average).

In the comparison between age readings from broken-and-burnt vs. thin-sectioned otoliths (Fig. 2.1.10c), Norway and Iceland had good correspondence between both preparation me-thods, while Canada and Poland assigned higher ages to thin-sectioned otoliths.


Table 2.1.10a Age reading results for exchange set E1 (pelagic S. mentella from the Norwegian Sea), break-and burn.

Fish information Age reading estimates (years) by nation1, confidence index (CI)

and average age Precision statistics Fish No


Avg. age PA CV APE

1 34 M 11 F 13 18 13 (13) 14 0% 22% 16%

2 40 F 39 FG 38 29 21 (17) 32 0% 27% 21%

3 39 F 28 FP 25 24 21 (17) 25 25% 12% 8%

4 38 M 41 FP 31 34 20 (17) 32 0% 28% 19%

5 37 F 18 F 20 18 21 (13) 19 0% 8% 7%

6 43 F 33 FP 35 33 26 (20) 32 0% 12% 9%

7 36 M 40 F 35 33 20 (15) 32 0% 27% 19%

8 32 M 19 F 19 16 13 (12) 17 0% 17% 13%

9 35 F 19 F 20 21 16 (13) 19 25% 11% 8%

10 33 M 19 F 19 19 15 (11) 18 0% 11% 8%

11 36 M 17 FP 19 18 16 (14) 18 25% 7% 6%

12 36 M 34 FP 33 19 21 (16) 27 0% 29% 25%

13 32 M 16 F 18 18 13 (12) 16 25% 15% 11%

14 40 M 38 FP 39 31 26 (17) 34 0% 18% 15%

15 32 M 18 F 19 17 14 (12) 17 25% 13% 9%

16 36 M 13 FP 18 20 17 (15) 17 25% 17% 12%

17 33 F 18 FG 20 17 13 (12) 17 25% 17% 12%

18 37 M 28 FP 25 27 20 (15) 25 25% 14% 10%

19 35 M 18 F 20 19 20 (13) 19 25% 5% 4%

20 38 M 35 FP 30 29 23 (18) 29 25% 17% 11%

21 39 F 36 FP 35 27 24 (18) 31 0% 19% 16%

22 28 M 8 F 10 9 11 (10) 10 25% 14% 10%

23 27 M 10 F 11 11 12 (9) 11 50% 7% 5%

Average 14.1% 16.0% 11.8%



Table 2.1.10b Age reading results for exchange set E1 (pelagic S. mentella from the Norwegian Sea), thin-sections.

Fish information Age reading estimates (years) by nation1, confidence index

(CI) and average age Precision statistics Fish No


Avg. age PA CV APE

1 34 M 20 P 15 17 13 (15) 16 0% 18% 14%

2 40 F 40 P 37 32 27 (18) 34 0% 17% 13%

3 39 F 40 P 26 23 29 (19) 30 0% 25% 18%

4 38 M 36 P 30 33 26 (17) 31 0% 14% 10%

5 37 F 20 P 20 18 25 (15) 21 0% 14% 11%

6 43 F 31 P 30 32 28 (19) 30 25% 6% 4%

7 36 M 42 P 37 35 25 (16) 35 25% 21% 14%

8 32 M 18 FP 19 18 17 (14) 18 50% 5% 3%

9 35 F 23 FP 20 21 22 (14) 22 25% 6% 5%

10 33 M 23 P 18 18 16 (12) 19 0% 16% 12%

11 36 M 19 FP 19 19 17 (14) 19 75% 5% 3%

12 36 M 26 P 27 19 21 (16) 23 0% 17% 14%

13 32 M 26 P 18 17 16 (12) 19 0% 24% 17%

14 40 M 40 P 37 32 27 (17) 34 0% 17% 13%

15 32 M 23 FP 20 18 18 (12) 20 25% 12% 9%

16 36 M 33 P 23 22 18 (16) 24 0% 27% 19%

17 33 F 27 P 20 17 14 (13) 20 25% 29% 20%

18 37 M 32 P 28 28 24 (16) 28 50% 12% 7%

19 35 M 26 P 23 19 19 (14) 22 0% 16% 13%

20 38 M 35 P 30 29 23 (17) 29 25% 17% 11%

21 39 F 35 P 37 26 25 (18) 31 0% 20% 17%

22 28 M 10 FP 11 10 10 (11) 10 75% 5% 3%

23 27 M 10 FP 11 11 11 (9) 11 75% 5% 2%

Average 20.7% 15.0% 10.9%



Figure 2.1.10a Age-bias plots for exchange set E1 (pelagic S. mentella from the Norwegian Sea, read by the break-and-burn method) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given a verage age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

Figure 2.1.10b Age-bias plots for exchange set E1 (pelagic S. mentella from the Norwegian Sea, read by the thin-sectioning method) by nation. Each error bar represents 2 standard deviations around the mean age assigned by a reader for all fish assigned a given a verage age (rounded to whole numbers) by all readers. The 1:1 equivalence line is also indicated.

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Canada

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Iceland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Norway

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Poland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Canada

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Iceland

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Norway

0

10

20

30

40

50

0 10 20 30 40 50

Mea

n ag

e +/

-2sd


Poland


Figure 2.1.10c Exchange set E1 (pelagic S. mentella from the Norwegian Sea). Comparison of age readings between break-and-burn and thin-sectioning methods by nation.

2.1.11 Juvenile Sebastes sp. from the Flem ish Cap

Exchange set E2 contained Sebastes sp. otoliths from the Flemish Cap of 8-17 cm length, col-lected by Spain on a research survey in 2007 (Table 2.1).

Age estimates ranged from 1 to 5 years (Table 2.1.11). The Canadian reader mostly had ‘fair’ or ‘fairly poor’ confidence in the readings, whereas the Spanish reader 1 (most familiar with this stock) had ‘fair’ to ‘good’ confidence. The PA values ranged from 14 to 43% (average 31.1%), the average CV was 34.8% and the APE was 25.1% (Table 2.1.11).

No age-bias plot was produced due to the low number of individuals in this set.

Table 2.1.11 Age reading results for exchange set E2 (juvenile Sebastes sp. from the Flemish Cap).

Fish No


Spain 1 CI

Spain 2

Avg. age PA CV APE

43_18 8 I 1 FP 1 2 2 1 F 3 2 29% 49% 33%

43_15 9 I 1 FP 1 2 3 1 F 3 2 14% 54% 42%

43_14 10 I 1 FP 1 2 3 1 F 3 2 14% 54% 42%

43_1 11 I 2 F 2 3 3 2 FG 3 3 43% 22% 17%

43_3 12 I 2 F 2 3 4 2 FG 3 3 29% 31% 22%

43_4 13 I 3 F 2 3 4 2 FG 3 3 43% 27% 17%

43_7 14 I 2 F 2 3 4 3 FG 3 3 43% 27% 17%

43_8 15 I 3 F 2 3 5 3 40% 39% 25%

43_10 16 I 3 F 3 3 4 3 G 5 4 14% 24% 21%

43_11 17 I 3 G 2 3 4 3 G 4 3 43% 24% 17%

Average 31.1% 34.8% 25.1%

0

10

20

30

40

50

0 10 20 30 40 50

Age

Thi

n-se

ctio

ns (y

ears

)

Age Break-and-burn (years)

Canada

0

10

20

30

40

50

0 10 20 30 40 50

Age

Thi

n-se

ctio

ns (y

ears

)


Iceland

0

10

20

30

40

50

0 10 20 30 40 50

Age

Thi

n-se

ctio

ns (y

ears

)


Norway

0

10

20

30

40

50

0 10 20 30 40 50A

ge T

hin-

sect

ions

(yea

rs)


Poland


3 Sources of age determination error in terms of bias and precision, interpretational differences between readers, ageing criteria

Considering the results of the exchange exercise (section 2.1), an analysis of potential sources of age determination error was performed in two manners:

1. Seven redfish readers, out of eight that participated in the exchange, were present in the work-shop (only Russia was missing) and discussed the ageing criteria in plenary.

2. A collection of 20 digital images of S. marinus and S. mentella from 3 to 39 years were se-lected for discussion, in plenary, on differences and discrepancies on criteria among readers.

The 2006 workshop (ICES 2006) revealed and documented specific age determination criteria where participants differed and recommended adoption of standards to promote compara-bility. These standards were used to identify if readers had adopted the standards and identi-fy which, potentially, were still the cause of disagreements produced from the 2007-2008 exchange. In particular, the discussion was focused on the interpretation of first annulus, the preferred reading axis and the interpretation of annuli close to the edge, following the main recommendation of the 2006 workshop:

The reader should avoid ageing along the otolith cross-section’s dorsal-distal axis and ALWAYS count annuli along a nucleus to proximal edge axis to avoid i) missing post maturity growth not visible on the dorsal axis and ii) mis-interpreting prominent checks as annuli. Ref. Protocol for redfish age de-termination, ICES CM1996/G:1 and MacLellan (1997). Agencies should formally document the date that they began to use this criteria consistently.

There were no major differences in opinion between the workshop members as to the prin-ciples and reading axes to use for interpreting the redfish otoliths provided. The larger dis-crepancies observed during the exchange were sorted out during subsequent discussions, and all age readers agreed to the same general interpretation rules. However, it was also agreed that there may be some uncertainty of interpretation of intermediate zones, approx-imately between ages 5 and 10. This may be looked at as a transition range between the first few large growth zones and the outer range of nearly equally spaced thin zones. In this tran-sition range, the annuli may appear in pairs, possibly indicating that two zones should be considered as one year of growth. The present practice is to count each zone, which could possibly lead to overestimation of age by 3-4 years.

Several of the participants felt that it may be useful to consider the size of the first annulus in order to separate it from checks. This may be especially important for younger individuals, where the interpretation could not be guided by the pattern of succeeding growth zones. A simple and effective way to ascertain correct identification of the first annulus is to use a measuring stick with marks corresponding to the approximate length of the first zone. This is quicker to use than measuring oculars, which have to be calibrated for each magnification. When referring to expected size of the first annulus, it is important to consider if the section was accurately placed in the centre of the core.

Some participants warned that even for younger fish, the number of checks along the longest axis may be confusing, and that this axis should be abandoned. It was a general feeling that a safe interpretation of the first several annuli might often be found along the medial side of the longest axis, where checks converge into more distinct zones.

Most of the differences seemed to be related with the fact that the otoliths exchanged were prepared with different methods (broken and burnt, broken and baked and thin-sectioned), since readers have experience with only one method and hence the relative inexperience with the method may produce inconsistent readings. Additionally, there are differences in the opt-


ical adjustments made on stereomicroscopes, depending on the method, that may have af-fected observed divergences in age readings.

Nevertheless, it was clear among readers that the quality of the otoliths is a crucial aspect to be considered when interpreting the age and when comparing readings. Quality means rea-dability in terms of the otolith natural pattern, but also the preparation of the otolith (break-ing, sectioning, burning). Among the sets of otoliths exchanged (see section 2.1), the lowest PA was obtained for set 6 (Table 3.1) where 80% of the material were assigned ‘fairly poor’ or ‘poor’ by the most experienced reader. This set was considered by all readers as having very poor quality and in general very difficult to read. Set 7 showed the second lowest PA and the highest CV (Table 3.1); this set was prepared originally with the break-and-bake method only used in Spain and only this country is familiar with this stock (Flemish Cap), as consequence in 80% of the cases, the Canadian reader had ‘fair’ or ‘fairly poor’ confidence in the readings, whereas the Spanish reader had ‘good’ or ‘fairly good’ confidence in 60% of the cases. This illustrate the importance of the method and stock on observed differences. Consensus existed about agreeing on the selection of a unique method to read otoliths as a routine, but specially when exchanging material for comparisons of criteria (see section 4).

Table 3.1 Summary results of the redfish otolith age readings from the exchange exercise.

SET P A CV AP E MEAN MIN MAX

1 34.20% 19.30% 13.70% 7.2 3 11

2 17.30% 17.50% 12.90% 24.9 9 42

3 34.60% 20.30% 14.70% 11.4 2 32

4 33.30% 15.3 14 17

5 17.60% 20.60% 15.00% 17.5 7 40

6 8.70% 21.70% 15.70% 25.9 16 31

7 11.70% 25.60% 20.10% 10.7 5 17

8 29.20% 19.20% 13.80% 9.8 5 17

9 20.10% 19.90% 14.90% 11.4 7 16

E1 14.10% 16.00% 11.80% 22.2 10 34

E1 20.70% 15.00% 10.90% 23.7 10 35

E2 31.10% 34.80% 25.10% 2.8 2 4

0%

5%

10%

15%

20%

25%

30%

35%

40%

0 5 10 15 20 25 30

Perc

ent a

gree

men

t

Mean age

Figure 3.1. Relationship between mean age and Percent of agreement for each exchanged set.

The longevity of redfish also influences on discrepancies reported. Readers agreed that due to the slow growth of redfish, annuli accumulate only in certain zones of the otolith and rea-dability becomes more difficult. In fact there was a significant negative correlation between the mean age of the set and the PA (Figure 3.1) when excluding the poor quality set (r2=0.63, F= 15.3, p<0.005). In this sense, it was agreed that the QA/QC procedure implemented in some labs and the confidence index used are useful tools for further discussion on sources of errors in age determination. During the 2006 workshop (ICES 2006), it was stated:

[…] The Workshop agreed that QC/QA should be implemented in all labs performing redfish age readings. QA is necessary to ensure good data quality, especially if age readings are going to be used in analytical assessment. The workshop also recommends developing for redfish the specific proce-dures used as Quality Control to test quality against set standards and the action to be taken when results don’t meet standards. As first step, a reading confidence index should be tested within otolith exchanges, agreed in the near future and implemented. […]

However, few labs have implemented the use of confidence index since then, and only two readers reported confidence indices routinely during the exchange (see section 6). The CI is


important to reflect not only readability in terms of the otolith pattern but also due to low quality equipment or to a non-adequate setup of the equipment in terms of optics and light.

In addition to different preparation methods, there were some differences between labs in procedures used when working with age estimation, e.g. relating to how to cut/break the oto-liths and to the access to biological information. In some labs otoliths are stored dry in paper envelopes while others store the otoliths in a liquid to enhance contrasts. This enhanced con-trast allows the reader to see the first growth zone from the surface of the otoliths and there-by securing a more precise placement of the breaking/sectioning. Some labs write biological information (fish length, sex, maturity) on the paper envelopes. Although most readers will not use this length information actively during interpretation, this practice may potentially reduce the quality of the ageing estimates.

These above discussions were accompanied by hands-on exercises on how to prepare and read otoliths (see section 4) to reach an agreement on this matter and proceed with age read-ings during the workshop attempting as much as possible to reduce methodological bias. During the exercises, the importance on using standard high-quality binocular/dissecting microscope equipment was highlighted, with large magnification (100x minimum, 200x rec-ommended) with sufficient clarity, using 20x oculars. Light should be sufficient, preferably using fibre optics necessary for higher magnifications and viewing of very small annual zones on the cross-sections of older otoliths (>30 yrs). Standard quality equipment promotes reproducibility between readers and agencies.

To analyse in detail the sources of age determination error in terms of bias and precision and the interpretational differences between readers, two sets of otoliths were used for compara-tive age reading during the workshop (Table 3.2). The first set of otoliths (Set A) was a selec-tion of exchanged otoliths (see section 2.1). Considering the limited time, only a fraction of the exchanged material could be re-read during the workshop. Thus, 40 otoliths were se-lected (Table 3.2) based on the following criteria:

1. All sets of otoliths exchanged should be represented with at least three otoliths, with the ex-ception of set 1 and 2 (not available at the workshop) and set 5 (due to the low quality of the otoliths); covering, thus, most of the redfish stocks in the North Atlantic.

2. Some of the otoliths where the agreement was 0% should be included with a good representa-tion of low to high ages. A total of 17 otoliths were selected using this criterion.

3. Otoliths where a reduced proportion of agreement was due to only one or two readers (present during the workshop). When excluding these readers, the PA was high. 18 otoliths, ranging from 14% to 50% agreement, were selected using this criterion.

4. Otoliths where the agreement was 100%. Five otoliths were selected using this criterion.

The second set of otoliths (Set B) was a selection of 20 otoliths made by PBS-DFO Fish Age-ing Laboratory among those exchanged during 2007-2008 outside the workshop otolith ex-change material (Table 3.2). These otoliths were selected among S. marinus and S. mentella from Iceland, Greenland and Irminger Sea used in radiometric age validation (Stransky et al. 2005b) and from Flemish Cap S. mentella. Samples of both burnt and thin-sectioned otoliths were chosen for clarity of pattern from a range of ages to best demonstrate application of cri-teria to workshop participants.

Workshop readers had various levels of experience and aged the samples independently of each other. The catch month of the samples was available to the readers, whereas individual fish information, such as fish length, was not revealed to prevent bias. Binocular microscopes were used, with reflected fibre optic lighting. A discussion binocular was set up with a cam-era system and PC connection to allow the group to follow the reading process on a comput-er screen after all readers having finished the exercise.


All material was prepared broken-and-burnt following the protocol described by MacLellan (1997). The ageing criteria that were used are described in detail in the previous workshop report (ICES 2006). For the time being, no alterations to these criteria have been recommend-ed.

Table 3.2 Redfish otolith samples used for comparisons between readers during the workshop.

DISCUSS ION SET

EXC HAN GED SET

SPEC IE S SAMPL ING AR EA (SUBAR EA OR

DIVISIO N)

COUNTRY SAMPL ING DATE O R

PERIO D

N

A 3 S. marinus Ice lan d (Va) Ice land March 2000 6

A 4 S. marinus Ice lan d (Va) Ice land October 1995 6

A 6 S. mentella Norwegian Sea (II) Russia August 2006 4

A 7 S. marinus Flemish Cap (NAFO 3M) Spain July 2007 9

A 8 S. mentella Flemish Cap (NAFO 3M) Spain July 2007 4

A 9 S. fascitus Flemish Cap (NAFO 3M) Spain July 2007 4

A E1 S. mentella Norwegian Sea (II) Norway September 2007 4

A E2 S. spp Flemish Cap (NAFO 3M) Spain July 2007 3

B New S. marinus Ice lan d (Va) Ice land March 2000 10

B New S. mentella East Greenland (ICES XIVb)

Germany Sept-Oct 2000 3

B New S. mentella Irminger Sea (ICES XII; NAFO 1F/2H/2J)

Germany June-July 2001 5

B New S. mentella Flemish Cap (NAFO 3M) Spain July 2007 2

3.1 Results for otolith set A Set A contained otoliths from a wide variety of ages previously read during the exchange (Table 3.1) from several areas and three different species.

Age estimates varied between 1 and 40 years (Table 3.3). The percent agreement (PA) values ranged between 11-89% (average 37.2%), the average coefficient of variation (CV) was 23.3% and the average percent error (APE) was 21.1% (Table 3.3). These results showed a notable improvement in PA when compared with the readings of the same readers of the same oto-liths made during the exchange (Section 2.1), as the agreement increased from 21 to 37%, but CV was at the same level (Table 3.3).

The age-bias plots (Figure 3.2) showed a slight non-linear bias for some readers. The regres-sion of age readings by each reader against modal age showed a high r2 for all readers (Table 3.4), with values of readers 3, 4, 7, 8 and 9 being the lowest. For readers 4 and 7, these values were the consequence of 1 or 2 otoliths considerably underestimated. However, a clear un-der-estimation of age by readers 3, 8 and 9 and a slight relative overestimation by reader 6 was observed. These are among the less experienced readers. If only the readings of the five most experienced readers are used, the PA increases to 42.5% and CV and APE decrease not-ably (18% and 15% respectively). Most of the variation consistently occurred among higher ages, i.e. >25 yrs; however, removing the otoliths with modal ages above 15 yrs improved only slightly the results (PA 45.5%). Individually, all readers improved the readings (com-pared through regression with modal ages) from the exchange to those read during the workshop.

Readers agree that the discussions during the workshop preceding these readings and espe-cially the training on methods (break-and-burn) and equipment setup have facilitated consi-derably the readings and potentially the greater agreement. However, it was agreed that a


major obstacle for a better agreement was the quality of the otoliths itself, and not only the technique.

Table 3.3: Workshop Set A. Age reading results for Set A read during the workshop.

NO

FISH SIZE (CM ) SEX

READ ER PR ECISION STA TISTIC S

1 2 3 4 5 6 7 8 9 MOD E PA CV APE 1 6 6 6 7 6 6 7 5 7 6 56% 11% 7% 2 4 4 8 4 6 4 6 5 5 4 44% 27% 28% 3 15 15 17 17 15 16 13 14 14 15 33% 9% 7% 4 2 3 3 1 2 2 3 4 2 2 44% 36% 33% 5 20 18 15 24 22 26 17 20 23 20 22% 17% 14% 6 23 28 18 27 26 24 27 20 20 27 22% 15% 13% 7 15 15 15 16 15 15 15 15 15 15 89% 2% 1% 8 15 14 17 15 15 18 15 13 17 15 44% 10% 7% 9 14 14 20 12 15 15 20 14 15 14 33% 18% 13%

10 19 22 21 27 18 24 22 14 21 22 22% 18% 12% 11 15 14 16 16 15 12 12 15 13 15 33% 11% 8% 12 14 15 13 16 14 15 11 14 12 14 33% 11% 7% 13 35 30 14 10 30 41 23 23 22 30 22% 39% 28% 14 40 36 13 34 20 40 24 22 24 40 22% 34% 28% 15 35 35 27 34 32 42 27 23 29 35 22% 18% 14% 16 33 36 24 28 31 37 38 25 29 31 11% 16% 15% 17 18 18 18 17 17 17 14 17 17 17 56% 7% 4% 18 18 17 17 19 17 15 12 16 16 17 33% 12% 8% 19 8 7 12 11 6 8 13 10 20 8 22% 40% 40% 20 4 4 9 6 3 5 5 7 3 4 22% 38% 39% 21 8 7 8 8 7 7 8 9 8 8 56% 9% 6% 22 7 9 11 12 8 4 7 14 7 7 33% 35% 35% 23 7 8 7 7 7 5 7 10 8 7 56% 18% 11% 24 8 4 8 8 6 5 7 9 12 8 33% 32% 21% 25 8 8 10 15 11 10 10 15 11 10 33% 24% 18% 26 9 10 11 12 10 11 13 19 12 10 22% 25% 21% 27 8 8 11 9 9 9 9 17 12 9 44% 28% 19% 28 4 7 7 6 4 6 5 10 4 4 33% 33% 47% 29 8 6 10 11 6 15 12 9 4 6 22% 38% 60% 30 8 8 6 9 8 20 11 9 8 8 44% 42% 30% 31 7 7 7 9 8 9 8 6 7 7 44% 13% 13% 32 13 14 11 14 10 17 14 12 13 14 33% 15% 12% 33 6 5 10 4 5 4 6 4 5 5 33% 34% 25% 34 12 11 12 14 13 12 13 11 14 12 33% 9% 6% 35 39 39 31 36 40 36 26 28 36 36 33% 15% 11% 36 36 33 20 32 36 37 32 28 28 36 22% 17% 13% 37 20 18 19 20 19 17 17 19 19 19 44% 6% 4% 38 2 3 4 2 2 2 2 3 2 2 67% 30% 22% 39 1 2 3 1 1 2 1 2 1 1 56% 47% 56% 40 1 2 5 1 1 1 2 3 1 1 56% 72% 89%

MEAN VAL UE S FO R READIN G S D URI NG W ORKSHOP (A LL )

37.2% 23.3% 21.1% MEAN VAL UE S FO R READIN G S D URI NG EX CHA NGE (A LL

20.9% 23.2% 27.1%

MEAN VAL UE S FO R READIN G S D URI NG W ORKSHOP (FIV E REA DE RS)

42.5% 17.9% 15.3%


Table 3.4: Regression parameters for age readings of Set A by each reader against modal age.

Reader a b r2 Level of expe-

rie nce 1 1.024 -0.238 0.984 24 yrs

2 0.984 0.203 0.975 20 yrs

3 0.500 5.842 0.674 <1 yr

4 0.835 2.557 0.859 19 yrs

5 0.881 1.044 0.895 10 yrs

6 1.069 0.280 0.915 0.5 yrs

7 0.744 2.914 0.853 14 yrs

8 0.591 5.043 0.818 10 yrs

9 0.754 2.825 0.863 1 yr

Figure 3.2: Age-bias plots of the selected Set A read during the workshop. The mean ages (in years) read by a certain reader are plotted against the modal age over readings of all readers. The error bars around the mean ages represent ± 2 standard deviations. The 1:1 equivalence is given as a straight line.


3.2 Results for otoliths set B Set B contained otoliths from a wide variety of ages (Table 3.5) previously read at PBS-DFO and selected by their extraordinary quality of the reading pattern. This col-lection was read by nine readers (reader 2 to 10) with different level of experience (Table 3.6).

Age estimates varied between 2 and 40 years (Table 3.5). The percent agreement (PA) values ranged between 0-100% (average 46.7%), the average coefficient of variation (CV) was 15.5% and the average percent error (APE) was 13% (Table 3.5). These re-sults showed a notable improvement in PA when compared with the readings of Set A (Section 3.1), as the agreement increased from 37 to 47%, and a reduction in CV from 23% to 15% was observed.

The age-bias plots (Fig. 3.3) showed a good agreement among readers, with no clear bias except for reader 9, having only short experience. The regression of age readings by each reader against modal age showed a high r2 for all readers (Table 3.6), with the lowest values observed for reader 9. If only the readings of the five most experienced readers are used, the PA increases to 57%, CV and APE decrease notably (9% and 8%, respectively). Again, most of the variation consistently occurred among higher ages, i.e. >25 yrs; however, the deviation was considerably lower than in previous exercise with otoliths of standard quality (Fig. 3.2).

Individually, all readers improved the agreement compared with modal age regard-ing previous sets of otoliths compared (exchange and set A). Readers agree that the method, equipment setup and the quality of the otoliths itself explain most of the ob-served differences and the remaining variation is due to slight differences in the in-terpretation of otolith pattern related with the fact of using different species and areas from those the reader is use to. Small exchanges among readers working in the same area and species and including most experienced readers may overcome these differ-ences.


Table 3.5: Workshop Set B. Age reading results for Set B read during the workshop.

NO

FISH SIZE (CM ) SEX

READ ER PR ECISION STA TISTIC S

2 3 4 5 6 7 8 9 10 MOD E PA CV APE 1 36 35 36 36 36 37 29 33 36 36 56% 7% 4% 2 15 15 14 15 13 14 14 15 15 15 56% 5% 4% 3 3 2 4 5 5 8 4 3 3 3 33% 43% 44% 4 18 18 18 19 15 19 19 14 19 19 44% 11% 7% 5 10 10 10 10 10 10 10 10 10 10 100% 0% 0% 6 20 22 21 20 20 20 21 14 20 20 56% 12% 6% 7 12 12 12 12 11 12 12 12 12 12 89% 3% 1% 8 6 9 6 6 6 8 6 6 7 6 67% 17% 11% 9 35 34 34 34 31 30 28 27 34 34 44% 9% 7%

10 26 26 26 27 23 27 25 25 26 26 44% 5% 3% 11 8 8 9 9 7 9 10 8 8 8 44% 10% 8% 12 6 9 8 7 8 8 10 5 6 8 33% 21% 15% 13 11 11 13 11 11 14 19 13 10 11 44% 22% 16% 14 55 50 56 58 55 53 42 39 57 55 22% 13% 8% 15 24 25 27 25 24 28 24 25 25 25 44% 6% 4% 16 70 49 57 64 48 59 47 27 67 56 0% 24% 19% 17 36 34 36 37 32 34 25 21 36 36 33% 17% 11% 18 31 31 30 31 27 30 31 12 31 31 56% 22% 9% 19 4 10 5 4 3 9 13 6 4 4 33% 53% 67% 20 18 20 18 18 16 19 21 16 17 18 33% 9% 7%

MEAN VAL UE S FO R READIN G S D URI NG W ORKSHOP (A LL REA DE RS)

46.7% 15.5% 13%

MEAN VAL UE S FO R READIN G S D URI NG W ORKSHOP (FIV E

57.0% 9.0% 8%

Table 3.6: Regression parameters for age readings of Set B by each reader against modal age.

Reader a b r2 Level of expe-

rie nce 2 1.123 -2.006 0.968 20 yrs

3 0.883 2.461 0.986 <1 yr

4 1.014 0.140 0.996 19 yrs

5 1.090 -1.086 0.988 10 yrs

6 0.931 -0.022 0.987 0.5 yrs

7 0.952 1.885 0.976 14 yrs

8 0.707 5.260 0.926 10 yrs

9 0.574 4.183 0.793 1 yr

10 1.114 -1.849 0.980 30 yrs


Figure 3.3: Age-bias plots of the selected Set B read during the workshop. The mean ages (in years) read by a certain reader are plotted against the modal age over readings of all readers. The error bars around the mean ages represent ± 2 standard deviations. The 1:1 equivalence is given as a straight line.

3.3 Conclusions

The workshop showed significant progress in the correspondence between readers compared with previous workshop results. Also there was a clear improvement in agreement between the exchange exercise and the readings during the workshop af-ter discussions, especially when using only clear-pattern otoliths and comparing the most experienced readers. Good quality otoliths with good readability showed only small differences among readers and demonstrated that in general there are no dif-ferences in age reading criteria among readers. This fact highlighted the importance of using a confidence index for each reading (see section 6). However, only three readers reported a confidence index for each otolith read during the exchange exer-cise, and the workshop encourages to continue the implementation of QA/QC in the different laboratories involved in age reading of redfish.

Only otoliths with good or fairly good confidence index should be used for compari-sons and discussion on ageing criteria among readers. However, it remained unclear if only those otoliths should be used for production ageing, especially for stock as-sessment purposes, as these clear-pattern otoliths may correspond to given ages or


cohorts and hence may bias the age-length key and the stock age composition if only those otoliths are used. More research in this direction is needed.

Additionally, it was also clear that several of the observed differences are considera-bly reduced when using a high-quality binocular/dissecting microscope equipment, especially when prepared for large magnification (100x minimum, 200x recommend-ed) with sufficient clarity, using 20x oculars. Light is also a key component especially for old individuals. It should not only be sufficient, but at higher magnifications, when viewing of very small annual zones on the cross-sections of older otoliths (>30 yrs), light should be of high quality, for which using fiber optics is recommended. Standardization of the equipment, and its quality, promotes reproducibility between readers and agencies.

Nevertheless, some differences for certain criteria were also detected, especially in:

i. The position of the first few annuli. Although discrepancies could be due to species/stock-specific differences, it was agreed that each country should measure the first three annuli in 10 otoliths with very clear patterns for a num-ber of stocks, and containing several year classes, following specific guidelines given in the report. To achieve this, each annulus should be measured as the distance across the nucleus from the dorsal tip to the ventral tip of the annulus. Measurements can be taken with a micrometer in one of the oculars, or more commonly over a calibrated digital picture.

The workshop recommends to create a reference collection within each lab for their specific stocks, and their pictures should be transferred to an image col-lection to be uploaded on a webpage with sufficient metadata (technical and biological information).

ii. The identification of several checks, especially during the first 3-4 years, and the interpretation of the transition zones. These differences in interpretation very likely are derived from each reader being familiar with some species or stocks. To overcome this problem, it is recommended to continue with the harmonization of the age reading within and across stocks and species follow-ing the guidelines, and through short-term exchanges among labs. To harmon-ize age reading within stocks, short-term exchanges should also be conducted among labs determining age of the same or related stocks.


4 Most appropriate otolith preparation method for age determina-tion of redfish and stock-specific guidelines for ageing

In the light of the above discussions and conclusions, it was agreed in plenary that for production ageing, the recommended otolith preparation method should be break-and-burn, following the guidelines described in Annex 4.

Several reasons were reported supporting this agreement:

• This is the routine method for the reference laboratory working on Sebastes species for more than 30 years: Pacific Biological Station, DFO, Nanaimo, Canada.

• This method has been adopted by most of the laboratories involved in redfish age determination.

• This is a relative quick, easy and inexpensive method compared with thin-sections, and still yielding very good results.

• Produces a more enhanced contrast than the break-and-bake method, reduc-ing potential bias or underestimation of age (see next section).

• Although burning requires some experience to produce a uniform darkening of the growth zones over the whole broken surface, unfamiliar readers got used to this method relatively quickly during the workshop, showing the versatility of the method.

Guidelines on how to determine age of redfish species are provided in Annex 4.

It was not possible to produce stock-specific guidelines for ageing during the work-shop. Participants agreed that preparation methods and general criteria on ageing do not differ greatly among stocks or even species. Potential differences may occur re-garding position of checks, prominent annuli, transition zones and width of the an-nual growth zones. However, these differences can be more related with cohort effects (checks) or species (transition zones) than stocks. Occasionally for some co-horts, stocks can be identified using these natural marks. For the above reasons, read-ers familiar with one stock may experience some minor problems ageing redfish from another stock, for example considering checks as annuli or vice versa. Harmonization of age reading within and across stocks and species should be continued following the guidelines, and through short-term exchanges among labs, especially among labs reading the same or related stocks.

There was a proposal for small exchanges for precision testing (see Section 6) that should be used also to harmonize readings within and across stocks.


5 Sex-specific growth information and differences in growth estimates between otolith readers and preparation methods

Sex information was available for most of the individuals used for age determination in the exchanged collection, i.e. 366 out of 381 otoliths, a total of 2,538 readings. Size at age was fitted to a von Bertalanffy growth curve for each sex separately (Fig. 5.1) disregarding reader, sex or stock, but considering only the break-and-burn method. Curves were compared using the Chow test (Chow, 1960; Saborido-Rey et al., 2004). There was no difference in growth among sexes (F=1.22, p>0.05).

All pair comparisons between readers showed significant differences in growth pat-tern (p<0.05). However, the plot of the growth curves by reader (Fig. 5.2) showed that generally growth curves followed a similar pattern, especially with a shift in growth rate between ages 10-15, except for readers from Poland and Russia, where shift in growth rate was less drastic. Excluding these readers, asymptotic length was very similar (40 cm). Most of the differences occurred in the longevity of the fish. Canada produced the older ages, and Russia the youngest. However, excluding again Russia and Poland, differences in longevity did not affect seriously the shapes of the curves, but only the maximum age, i.e. differences among readers occurred already in the asymptotic area of the curve (fish older than 25-30 years), and hence not affecting growth pattern. Very likely these differences have few, if any, impact on assessment. The Polish reader owned low experience in redfish, while Russian readers were more used to scale reading, in both cases affecting the results.

All reading methods (Break-and-burn, Break-and-bake and thin section) showed sig-nificant differences among them (p<0.05). Break-and-burn and thin sections yielded similar results, anyway, with a similar growth pattern, in spite of data lacking for small fish for the thin-sectioning method. However, break-and-bake clearly underes-timate age, showing a faster growth than the other two methods. The amount of read-ings is not sufficient to conduct a proper analysis, and additionally only one stock (Flemish Cap) and one country (Spain) read the exchanged otoliths using this me-thod, so the results may be biased. But it was agreed to select break-and-burn as the preferred method for production ageing.

Age (yr)0 10 20 30 40 50 60 70

Leng

th (c

m)

0

10

20

30

40

50

60

Females Males

Figure 5.1: Mean size at age and von Bertalanffy growth curves for males and females of the ex-changed collection of otoliths (only break and burn method considered).


Age (yr)0 10 20 30 40 50 60 70

Leng

th (c

m)

0

10

20

30

40

50

CanadaIcelandNorwayPolandRussia 1 Russia 2 Spain 1Spain 2

Figure 5.2: von Bertalanffy growth curves for each reader of the exchanged collection of otoliths (only break-and-burn method considered).

Age (yr)0 10 20 30 40 50 60 70

Leng

th (c

m)

0

10

20

30

40

50

60

Break and Burn Break and Baked Thin Sections

Figure 5.3: Mean size at age and von Bertalanffy growth curves for each method of the exchanged collection of otoliths.


6 Progress in the implementation of quality control and quality assurance (QC/QA) in redfish ageing labs Quality assurance is the system employed to ensure the quality of age data. It usual-ly involves a series/suite of quality control measures such as precision tests, use of indirect/direct assessments of accuracy and other standardized protocols such as training programs, documentation, age designation systems, standard age data sheet protocols, reference samples, etc.

For those interested in determining ages of North Atlantic redfish, it is necessary to continually compare readings and to resolve differences in interpretation when they occur. To achieve this, a number of routine steps should be followed, and testing sys-tems and standard procedures must be developed and documented. It is important that these are stock/species specific.

a) To participate in production ageing, a reader first must be able to demon-strate good “self-precision”, i.e. be consistent when independently ageing the same otoliths at different times. It is important to monitor for “reader drift” (change in interpretation) over time by periodic self tests or testing with a reference collection. Reference collections can take the form of actual samples or be a collection of images.

b) Between readers: readers of similar experience/ability should not demon-strate bias, and must exhibit precision. This can apply either between differ-ent readers in the same laboratory, or between readers in different laboratories.

c) Reference collections of both actual samples and images of otoliths should be established by each lab for the species/stocks they normally work with and be maintained to allow checking for reader "drift." Again it is emphasized that these must be stock/species specific.

d) Each laboratory should develop a routine bias/precision testing system that is made part of the agency's overall protocol. Some examples of procedures and analyses may be found in Anon. (1994), Campana et al. (1995) and Hoenig et al. (1995). Minimum precision testing requirements are often difficult to es-tablish and may have to be based on the impact it will have on production numbers. Turn-around time for data required may also have to be taken into consideration. Many labs precision test between 10-20% of the fish in each sample. Testing of each sample is encouraged as the reader becomes accus-tomed to this process as a “normal” part of producing good quality age data, to be looked on as an opportunity to maintain and improve expertise.

e) Regular inter-laboratory first hand discussion and comparison is neces-sary.This should apply most frequently to those involved with the same stock/species, and on a less frequent basis should involve all those involved with the same species regardless of stock (see http://care.psmfc.org

f) There is a need for continued discussion between the age readers and the bi-ologists familiar with the stock/species. The checking of outliers by data users should be part of standard QA/QC measures. They should develop a good

for full in-formation on exchange protocols). A common terminology and protocols should be followed by all laboratories involved in redfish ageing (see Annex-es 3 and 4). It was agreed to conduct small exchange exercises for precision testing (Table 6.1).


relationship with readers to provide feedback that can aid improvement of age data quality.

Table 6.1 Proposal for small exchanges for precision testing.

Species Stocks Countries Exchange

S. mentella Barents Sea-Norwegian Sea NOR, RUS, POL NOR, RUS, POL S. mentella Ice lan dic Sea - S. mentella Irminger Sea-Pe lagic POL ICE RUS (Scales) S. mentella Flemish Cap ESP CAN S. marinus Barents Sea-Norwegian Sea NOR, RUS ESP S. marinus Ice lan dic Sea ICE NOR S. marinus Flemish Cap ESP ICE S. fasciatus Flemish Cap ESP CAN

The participants agreed that it is essential to implement the use of confidence index indicate the reader’s confidence in assigning an age for redfish. Currently only three laboratories have implemented the use, in routine way, of the confidence index for redfish. This Quality Assurance (QA) system for production redfish age determina-tion is important to determine when reported differences are due to low quality (readability) otoliths, as well to select the collection for exchange and comparison and the reference collection (see item c) above).

The workshop recognizes that the Fish Ageing Lab at the Pacific Biological Station has been using a 5 level confidence index with success for more than 20 years, and hence decided to adopt such index (Table 6.2).

Table 6.2. Reader Confidence Index for Age Estimates used at Fish Aging Lab at Pacific Biological Station, Canada.

CONFIDENCE INDEX

ABBREVIATION QUALITATIVE MEANING (PATTERN CLARITY)

QUANTITATIVE MEANING (REPEATABILITY)

AGE EXAMPLE

S

Good G Pattern is very clear with no interpretation problems

Reader would always get the same age

10G, 57G

Fairly good FG Pattern is c lear with a few easy interpretation problems

Reader would get the same age most of the time

3FG, 58±

2FG

Fair F Pattern is fairly clear with some areas presenting easy & moderate interpretation problems

Reader would be within 1 yr all the time for fish <20 & 2-3 for fish > 30 yrs, etc

13±1F, 26±2F

Fairly poor FP Pattern is fairly unclear presenting a number of difficult interpretation problems

Reader would be within 2 yrs most of time for fish aged <20 & 3-5 yrs for fish >30yrs, etc

9±2FP, 63±5F

P

Poor P Pattern is very unclear presenting significant interpretation problems

Reader has little confidence in repeatability of age within 5-10 yrs or more in case of older fish

14±5P, 39±10

P


7 Publication of the results of the 2006 and 2008 workshops and the recent otolith exchanges (2007–2008) in the ICES CRR serie

As recommended in the 2006 workshop (ICES 2006) and endorsed by ICES PGCCDBS (Planning Group on Commercial Catches, Discards and Biological Sam-pling; ICES 2008), a publication of the results of the 2006 and 2008 workshops and the otolith exchanges 2007-2008 (section 2.1) in the ICES Cooperative Research Reports (CRR) series is envisaged and will be discussed at the PGCCDBS meeting in March 2010.

8 Recommendations

i. Acquire standard high-quality binocular/dissecting microscope equipment, especially obtain large magnification (100x minimum, 200x recommended) with sufficient clarity, using 20x oculars. Light should be sufficient, prefera-bly using fibre optics necessary for higher magnifications and viewing of very small annual zones on the cross-sections of older otoliths (>30 yrs). Standard quality equipment promotes reproducibility between readers and agencies.

ii. Each country should measure the first three annuli in 10 otoliths with very clear patterns for a number of stocks, and containing several year classes, fol-lowing specific guidelines given in the report.

iii. For production ageing, the recommended otolith preparation method should be break-and-burn, following the guidelines described in the report.

iv. All laboratories involved in redfish ageing should implement QA/QC proce-dures, mainly the use of a confidence index for each reading and the devel-opment of a routine bias/precision testing system.

v. Harmonization of the age reading across stocks and species should be con-tinued following the guidelines, and through short-term exchanges among labs.

vi. Create a reference collection within each lab for their specific stocks, and their pictures should be transferred to an image collection to be uploaded on a webpage with sufficient metadata (technical and biological information).

vii. For stock assessment and regular precision monitoring, the proposed confi-dence index should be used.


9 Literature cited

(including those cited in the annexes)

Anon. 1994. Report of the Workshop on Sampling Strategies for Age and Maturity. ICES CM 1994/D:1.

Andrews, A.H., Kerr, L.A., Cailliet, G.M. Brown, T.A., Lundstrom, C.C. and Stanley, R.D. 2007. Age validation of canary rockfish (Sebastes pinniger) using two independent otolith tech-niques: lead-radium and bomb radiocarbon dating. Mar. Freshw. Res., 58: 531–541.

Beamish, R.J., McFarlane, G.A. 1983. The forgotten requirement for age validation in fisheries biology. Trans. Am. Fish. Soc. 112(6): 735- 743.

Bedford, B.C. 1983. A method for pre paring sections of large numbers of otoliths e mbedded in black polyester resin. J. Cons. 41: 4-12.

Campana, S .E., Annand, M.C., and McMillan, J. I. 1995. Graphical and statistical methods for determining the consistency of age determinations. Trans. A m. Fish. Soc. 124: 131-138.

Chilton, D.E. and Beamish, R.J. 1982: Age determination methods for fishes studie d by the Groundfish Program at the Pacific Biological Station. Can. Spec. Publ. Fish. Aquat. Sci, 60: 102p

Chow, G.C. 1960. Tests of equality between sets of coefficients in two linear regressions. Eco-nometrica, 28: 591-605

Gaemers, P.A.M. 1984. Taxonomic position of the Cichlidae (Pisces, Perc iformes) as demon-strated by the morphology of their otoliths. J. Zool. 34: 566-595.

Harkonen, T. 1986. Guide to the otoliths of the bony fishes of the northeas t Atlantic . Danbiu Aps, Charlottenlund, Denmark.

Hoenig, J. M., Morgan, M. J., and Brown, C. A. 1995. Analysing differences between two age determination methods by tests of symmetry. Can. J. Fish. Aquat. Sci. 52: 364-368.

ICES. 1983. Report of the redfish ageing workshop, Bremerhaven, 14–18 Feb. 1983. ICES Doc-ument, CM 1983/G: 2. 2 pp.

ICES. 1984. Report of the workshop on ageing of redfish. ICES Document, CM 1984/G: 2. 9 pp.

ICES. 1991. Report of the workshop on age determination of redfish. ICES Document, CM 1991/G: 79. 18 pp.

ICES 1996. Report of the workshop on age reading of Sebastes spp. ICES C. M., 1996/G:1: 32 pp.

ICES 2006. Report of the Workshop on Age Determination of Redfish (WKADR). ICES CM 2006/RCM:09, 47 pp.

ICES 2008. Report of the Planning Group on Commercial Catches, Discards and Biological Sampling (PGCCDBS), 3‐7 March 2008, Cyprus. ICES CM 2008/ACOM:29, 91 pp.

Jensen, A.C. 1965. A standard terminology and notation for otolith age readers. ICNAF Res. Bull. 2: 5-7.

Kastelle, C.R., Kimura, D.K. and Goetz, B.J. 2008. Bomb radiocarbon are validation of Pacific ocean perch (Sebastes a lutus) using new statis tical methods. Can. J. Fish. Aquat. Sci. 65: 1101-1112.

Leaman, B.M. and Nagtegaal, D.A. 1987. Age validation and revised natural mortality rate for yellowtail rockfish. Trans. Am. Fish. Soc. 116: 171-175.

MacLellan, S . E. 1997. How to age rockfish (Sebastes) using S. a lutus as an e xample - the otolith burnt section technique. Can. Tech. Rep. Fish. Aquat. Sci., 2146: 42 pp.


Saborido-Rey, F., Garabana, D., and Cerviño, S . 2004. Age and growth of redfish (Sebastes mari-nus, S. mentella , and S. fasciatus) on the Flemish Cap (Northwest Atlantic) ICES Journal of Marine Science, 61: 231-242.

Secor, D.H., Dean, J.M. and Campana, S .E. 1995. Recent developments in fish otolith research. University of South Carolina Press. Columbia, USA, 735 pp.

Stransky, C., Gudmundsdottir, S ., S igurdsson, T., Lemvig, S., Nedreaas, K., and Saborido-Rey, F. 2005a. Age determination and growth of A tlantic redfish (Sebastes marinus and S. mentel-la ): bias and precision of age readers and otolith pre paration methods. ICES J. Mar. Sci., 62: 655-670.

Stransky, C., Kanisch, G., Krüger, A., and Purkl, S . 2005b. Radiome tric age validation of golden redfish (Sebastes marinus) and deep-sea redfish (S. mentella ) in the Northeast A tlantic . Fish. Res., 74: 186-197.

Wallace and Tagart 1994. Status of the coastal black rockfish stocks in Washington and north-ern Oregon in 1994. Appendix F. Pacific Fisheries Management Council, Portland, Orge-gon.

Williams, T. and B.C. Bedford. 1974. The use of otoliths for age determination. p.114-123. T.B. Bagenal [ed.] The ageing of fish. Proceedings of an International Symposium. Unwin Brothers Ltd., Surrey, England.

Wilson, C.W., Beamish, R.J., Brothers, K.D., Carlander, K.D., Casselman, J.M., Dean, J.M, Jearld Jr., A, Prince, E.D., and Wild, A. 1987. Glossary, p. 527-530. In R.C. Summerfelt, and G.E. Hall (eds.), Age and Growth of Fish, Iowa State University Press, Ames, Iowa.


Annex 1: List of par ti cipants

NAME ADDRESS PHONE/FAX EMAIL

Ole Thomas Albert

Institute of Marine Research P.O. Box 6404 N-9294 Tromsø Norway

TEL: +47 77 60 97 36 FAX: +47 77 60 97 01

[email protected]

Barbara Campbell

Fisheries & Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, B.C. V9T 6N7 Canada

TEL: +1 250 756-7179 FAX: +1 250 756-7053

[email protected]

Karen Charles Fisheries & Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, B.C. V9T 6N7 Canada

TEL: +1 250 756-7179 FAX: +1 250 756-7053

[email protected]

Nora Crosby Fisheries & Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, B.C. V9T 6N7 Canada

TEL: +1 250 756- FAX: +1 250 756-7053

[email protected]

Mariña Fabeiro Institute of Marine Research Eduardo Cabello 6 36208 Vigo Spain

TEL: +34 986 231930 FAX: +34 986 292762

[email protected]

Darlene Gillespie


TEL: +1 250 756-7121 FAX: +1 250 756-7053

[email protected]

Barbara Grabowska

Sea Fisheries Institute (MIR) ul. Kollataja 1 81-332 Gdynia Poland

TEL: +48 58 7356 206/274 FAX: +48 58 7356 110

[email protected]

Joanne Groot Fisheries & Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, B.C. V9T 6N7 Canada

TEL: +1 250 756- FAX: +1 250 756-7053

[email protected]

S if Guðmundsdóttir

Marine Research Institute P.O. Box 1390. Skúlagata 4 IS-l21 Reykjavík Iceland

TEL: +354 575 2000 FAX: +354 575 2001

[email protected]

Lise Heggebakken

Institute of Marine Research P.O. Box 6404 N-9294 Tromsø Norway

TEL: +47 77 60 97 26 FAX: +47 77 60 97 01

[email protected]


NAME ADDRESS PHONE/FAX EMAIL

Mary-Jane Hudson


TEL: +1 250 756-7179 FAX: +1 250 756-7053

[email protected]

Kristján Kristinsson

Marine Research Institute P.O. Box 1390. Skúlagata 4 IS-l21 Reykjavík Iceland

TEL: +354 575 2000 FAX: +354 575 2001

[email protected]

Svend Lemvig Institute of Marine Research P.O. Box 1870 Nordnes, N-5817 Bergen Norway

TEL: +47 55 238689 FAX: +47 55 238687

[email protected]

Diana Little Fisheries & Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, B.C. V9T 6N7 Canada

TEL: +1 250 756-7179 FAX: +1 250 756-7053

[email protected]

Shayne MacLellan


TEL: +1 250 756-7179/-7189 FAX: +1 250 756-7053

[email protected]

Susan Mahannah


TEL: +1 250 756- FAX: +1 250 756-7053

[email protected]

Judy McArthur Fisheries & Oceans Canada Pacific Biological Station 3190 Hammond Bay Road Nanaimo, B.C. V9T 6N7 Canada

TEL: +1 250 756-7179 FAX: +1 250 756-7053

[email protected]

Fran Saborido-Rey (Co-chair)

Institute of Marine Research Eduardo Cabello 6 36208 Vigo Spain

TEL: +34 986 214466 FAX: +34 986 292762

[email protected]

Arne Storaker Institute of Marine Research P.O. Box 1870 Nordnes, N-5817 Bergen Norway

TEL: +47 55 238688 FAX: +47 55 238687

[email protected]

Christoph Stransky (Co-chair)

Johann Heinrich von Thünen Institute [vTI] Federal Research Institute for Rural Areas, Forestry and Fisheries Institute of Sea Fisheries Palmaille 9 22767 Hamburg Germany

TEL: +49 40 38905 228 FAX: +49 40 38905 263

[email protected]


Margaret Treble Fisheries & Oceans Canada Freshwater Institute 501 University Cres. Winnipeg, Manitoba R3T 2N6 Canada

TEL: +1 204 984 0985 FAX: +1 204 984 2403

[email protected]

Kordian Trella Sea Fisheries Institute (MIR) ul. Kollataja 1 81-332 Gdynia Poland

TEL: +48 58 7356 266 FAX: +48 58 7356 110

[email protected]


Annex 2: Agenda

Tuesday, September 2

09:00-09:45 Opening of meeting (Chair: C. Stransky) Welcome by Dr. Laura Brown (Site Director and Head of the Division) Welcome (Shayne MacLellan), local arrangements, incl. computer and network arrangements

09:45-10:30 Introduction Adoption of Agenda Brief overview of ToRs, discussion Elections of rapporteurs

10:30-10:45 Break 10:45-11:15 Background review of past ICES WKADR work (F. Saborido-Rey/C.

Stransky) Brief presentations on methods and ageing experience of each laboratory.

11:15-12:00 ToR a: Introduction (C. Stransky) Updating and revision of 2006 report Presentation on otolith exchange comparative analyses

12:00-13:00 Lunch 13:00-13:30 Presentation on otolith exchange comparative analyses. Cont’d 13:30-15:00 Review of ageing criteria and methodology (S. MacLellan) 15:00-15:15 Break 15:15-16:00 Mini-workshop on Break-and-Burn technique (PBS staff) 16:00-16:15 ToR b: Introduction (F. Saborido-Rey) 16:15-17:15 Identification of potential sources of variation (C. Stransky) 17:15-17:30 Presentation of otolith selection for re-reading and discussion (F. Saborido-

Rey)

Wednesday, September 3

9:00-10:30 Age reading in plenary. Discussion on differences and discrepancies 10:30-10:45 Break 10:45-12:00 Age reading in plenary. Discussion on differences and discrepancies 12:00-13:00 Lunch 13:00-15:00 Age reading. 15:00-15:15 Break 15:00-15:30 Buehler otolith sectioning machine demo 15:30-17:30 Age reading

Thursday, September 4

9:00-10:30 Age reading. 10:30-10:45 Break


10:45-12:00 Revision of results from age readings 12:00-13:00 Lunch 13:00-15:00 Discussion on interpretational differences between readers and

laboratories. 15:00-15:15 Break 15:00-15:30 Leica scope demo 15:30-17:00 Agree on the ageing criteria. Manual for redfish age reading

Friday, September 5

9:00-10:30 ToR c: Introduction and presentation of results (F. Saborido-Rey) Discussions, Conclusions

10:30-10:45 Break 10:45-12:00 ToR d: Introduction and presentation of results (F. Saborido-Rey)

Discussions, Conclusions 12:00-13:00 Lunch 13:00-15:00 ToR e: Introduction (C. Stransky)

Brief presentations of the QC/QA progress in each lab Discussions, Conclusions

15:00-15:15 Break 15:15-16:30 ToR f: Introduction (C. Stransky)

Discussions, Conclusions 16:30-17:30 Final discussions and agreements. Recommendations


Annex 3: Glossary for age deter minat ion of red fish (updated from ICES 1996)

Much international work has taken place in order to standardize, as much as possi-ble, the terminology used during otolith interpretation. Participants reviewed exist-ing definitions (Beamish and McFarlane 1983; Chilton and Beamish 1982; Gaemers 1984; Harkonen 1986; Jensen 1965; MacLellan 1997; Secor et al. 1995; Wilson et al. 1987), and adapted them as appropriate for interpretation of otoliths from North At-lantic redfish. It is recommended that the following definitions be adhered to when making reference to North Atlantic redfish otoliths and their interpretation:

Accuracy: The closeness of a measured or computed value (e.g. age) to its true value. Accuracy can be proven or estimated: estimates of accuracy are less valuable, but in some cases only an estimate is possible.

Age estimation, age determination: These terms are preferred when discussing the process of assigning ages to fish. The term ageing (aging) should not be used as it refers to time-related processes and the alteration of an organism's composition, structure, and function over time.

Age-group: The group of fish that has a given age (e.g., the 5-year-old age-group). The term is not synonymous with year-class.

Annulus (pl. Annuli): (Winter zone) A translucent growth zone that forms once a year representing a time of slower growth. For most of the redfish stocks the annulus is formed during the fall and winter months. The translucent zone allows passage of light. In untreated otoliths under transmitted light, the translucent zone appears bright. Under reflected light it appears dark. When an otolith is burnt all translucent zones turn brown and fairly opaque.

Annual growth zone: A growth zone that consists of one opaque zone (summer zone) and one annulus (winter zone).

Bias: A lack of precision that is not normally distributed around the mean; it is skewed to one side or the other. For age determination it may apply to one reader's interpretations which are predominantly more or less than those of another for a por-tion of or all ages of the age range.

Checks: Translucent zone(s) that forms within the opaque (summer) zone, also representing a slowing of growth. These zones are usually not as prominent as annuli and should not be included in the age estimate. More than one check per year may form, especially in juvenile growth zones where they are most prominent.

Cohort: A group of fish that were born during the same calendar year.

Edge growth: The amount and type of growth (opaque or translucent) on an otolith’s margin or edge. Edge growth must be related to the time of year the fish was caught and the internationally accepted and standard January 1st birthdate. New opaque growth forming on the margin of a broken or sectioned otolith is often referred to as plus growth or increment growth.

Nucleus: It is the central area of the otolith formed during the larval stage. On Atlan-tic redfish otoliths the opaque central area is bound by the first translucent zone (check). Opaque zone: (Summer zone) A growth zone that restricts the passage of light. In untreated otoliths under transmitted light, the opaque zone appears dark.


Under reflected light it appears bright. Burning causes only a slight change in the co-lour of the opaque zone and does not change light transmission.

Precision: A process that measures the closeness of repeated independent age esti-mates. Precision relates to reproducibility and is not a measure of accuracy. The de-gree of agreement among readers is a measure of the precision of the determinations and not the accuracy of the technique.

Reflected light: Light that is shone onto the surface of an otolith from above, or from the side.

Sagitta (pl. sagittae): The largest of three otolith pairs found in the membranous la-byrinth of redfish. It is usually compressed laterally and is elliptical in shape. The sagitta is the otolith used most frequently in otolith studies.

Sulcus acusticus (commonly shortened to sulcus): A groove that forms along the proximal surface of the sagitta. A thickened portion of the otolithic membrane lies within the sulcus. This region is frequently referred to in otolith studies because of the clarity of growth zones near the sulcus in transverse sections of sagittae.

Summer zone: Opaque growth that is normally deposited during the spring and summer seasons when fish are growing relatively quickly.

Transition zone: A region of change in an otolith growth pattern between two similar or dissimilar regions. It is recognized as a region of significant change in the form (e.g., width or clarity) of the annual growth zones. In redfish, a transition zone has been defined as the region of change from juvenile to mature growth. The juvenile annual growth zones are relatively larger than those of later adult zones. For some fishes, this transition zone has been validated as coinciding with the onset of first ma-turity. This is also believed to be true for redfish, although not validated yet. Other areas of the broken otolith or section may also show a change in width and clarity of the annual growth zones which may be related to habitat changes (e.g., movement to deeper waters).

Translucent zone: (Hyaline zone, annulus, check) A growth zone that allows a better passage of light. The definition of the term hyaline has often been misunderstood and is not recommended for use.

Transmitted light: Light that is passed through the otolith from below (e.g., thin-sections); od from the side in the case of broken otoliths when the surface is sha-dowed.

Winter zone: Translucent growth (annulus; not check) that is normally deposited during the fall and winter seasons when fish are growing relatively slowly

Year-class: The cohort of fish that were born in a given calendar year (e.g., the 1990-year-class).

Zone: Region of similar structure or optical density (opaque or translucent). Syn-onymous with ring, band, and mark. The term zone is preferred.


Annex 4: Pro tocol for handl ing and age determinat ion of North At lant ic redf ish otoli ths

(updated from ICES 1996)

I: SAMPLING AND STORING

Both sagittae otoliths should be collected. In the North Atlantic, they are usually stored dry in a paper envelope, whereas Pacific Ocean redfish otoliths are routinely stored in glycerine. Both procedures are acceptable. Biological data such as total length, round weight, sex, maturity stage, etc. together with a reference number should be taken.

II: EQUIPMENT

Good quality equipment is essential towards attaining the most accurate and consis-tent age estimates possible. Microscope magnification, optics and lighting are the most important factors. In order to age otoliths, either broken, broken/burnt, bro-ken/baked or thin-sections, the reader needs a stereo/dissecting microscope (binocu-lar) with high quality optics and resolution power. It should be capable of magnifying at least 40x (experience from age determination of Pacific rockfishes would say at least 100-200x to correctly assess the age of older otoliths). These high magnifications, especially when reading broken and burnt otoliths, require very di-rect and intense light, and fibre optics are recommended.

Tissues are needed to wipe the otolith surface clean and dry in preparation for burn-ing or baking.

For broken/burnt otoliths, an alcohol lamp should be used to burn the otoliths. A pair of good quality, easily manipulated forceps with serrated tips is needed to hold the otolith during burning.

Paint brushes are used to paint oil (or glycerine/water) onto the broken otoliths (burnt, baked or unburnt). Mineral oil is recommended as it is nonreactive and non-toxic. A 50:50 mixture of alcohol and glycerine may also be used. A probe or forceps can be used to manipulate the position of the otolith in its plasticine or clay holder. A small dish can be used to hold the plasticine. Contrasting colours such as black or green plasticine are preferred.

III: PREPARATION TECHNIQUES

It is important to break or section the otolith as closely through the nucleus centre as possible (Figure A.1). Always look at the surface of the otolith half after breaking to be sure this is the case. Breaking or sectioning the otolith too far from the centre may cause the loss of the first annulus on the cross-section. A diagonal break or section will distort the whole pattern. It is important that the reader learn to recognize if these problems have occurred and adjust interpretation of the pattern accordingly. There is always the other otolith to work with. Burning causes translucent zones to turn brown. It is not necessary to burn both halves of a broken otolith if the first burnt half provides a clear pattern for a confident age determination.. The reader should assess the quality/clarity of the pattern regardless of preparation technique, and de-


termine if further preparations are necessary. A reader should read the same section 2-3 times before deciding on the most appropriate age assignment.

Figure A.1. Schematic drawing of a whole redfish otolith.

a) Breaking

The otoliths can be broken by the fingers, a tong, by sawing, scoring with a scalpel or saw blade etc. If using a saw, the blade should preferably be coated by fine diamond grit to prevent scratching of the cross-section surface. Some laboratories prefer the clarity of a "rough" broken surface to a polished or saw-cut surface.

Addition of oil to the cross-section surface is necessary to clarify the growth pattern. Mineral oil is recommended as it is inexpensive, non-toxic, and non-reactive In addi-tion, otoliths treated with it can be re-burnt or re-baked without leaving a dark resi-due adhering to the cross-section that may interfere with pattern interpretation.

i) Broken and Burnt Before breaking an otolith, wipe off all fluids, dirt or membranes using a tissue. Wa-ter in a dish can also be used to help remove organic tissues or dirt. Take special care to clean out the sulcus which often retains membranous materials. Fluids, oil or membranes that get onto the broken surface will burn black and obscure the growth zones.

Alcohol lamps are used to burn the broken otoliths. It is important to produce an "even" burn; that is, a uniform darkening of the growth zones over the whole broken surface. This can be difficult to achieve, but becomes easier and faster with expe-rience. Hold the otolith in forceps with the cross-section surface facing the eye so that it can be observed as it burns. It doesn't matter whether the sulcus is closest to or away from the flame. In general, the reader has more control over the speed and ex-tent of burning the further the otolith is held above the flame. The thinner parts (tips) and outer edges burn more quickly. To achieve an optimum burn, the otolith must be moved vertically or horizontally over the flame of the alcohol lamp to prevent por-tions from being burnt more/less than others. It can be removed from the flame to check progress either by eye or under the scope and then returned to the flame if more burning is required. Burning the otolith beyond dark brown will cause the oto-lith to turn black, grey and then ash white, all of which are considered over-burnt.


Different degrees of burning (light to dark) can be utilized to enhance certain charac-teristics on some broken areas of the otolith cross-section.

WARNING! Do not handle just-burnt otoliths with your fingers! They stay very hot for a few minutes and will burn flesh!

Generally, small or young otoliths should be burnt more lightly than bigger-thicker, older otoliths. Deliberately "over-burning" big, old otoliths can make older annual zones near the proximal edge clearer, and also eliminate checks. Paired otolith sam-pling provides readers with the opportunity to burn four halves when initial burns are not satisfactory. Experienced readers are less likely to need all four halves to produce a satisfactory estimate.

ii) Broken and Baked Before breaking an otolith, wipe or wash off all fluids, dirt or membranes using a tis-sue. Take special care to clean out the sulcus which often retains membranous mate-rials. Fluids, oil or membranes that get onto the broken surface will bake black and obscure the growth zones.

Otolith halves are then baked in an oven for at least one (1) hour at 200°C. This causes a light burn that is too light for being recommended for reading in reflected light. However, baking will enhance the surface of the otolith cross-section when the otolith is read in transmitted light. Baking tends to darken otolith growth zones in a more uniform manner over the whole broken surface, as compared to using an alcohol lamp (Saborido-Rey et al., 2004).

b) Thin Sections

To prepare large numbers of otolith thin sections the method described by Bedford (1983) is recommended. The otoliths are carefully positioned in rows in specially pre-pared moulds which are then filled with liquid black polyester resin. The resin har-dens to form solid rectangular blocks with the otoliths embedded in them. The blocks are removed from the moulds and machined with a high-speed diamond saw. Thin slices (0.6 mm thick) are cut from the blocks precisely along the lines of the centres of the rows of otoliths. The slices are mounted and fixed on standard glass microscope slides, and are then ready for age determination.

c) Digital pictures

Do some research, compare specs and chose a good quality digital camera. Digital microphotography colour cameras with a resolution ability of at least 3.5-5 megapix-els are recommended in order to obtain clear images of otolith cross-sections contain-ing small close together growth zones. Be sure to investigate the size of the light sensor (the larger it is the better the light gathering ability) as this will affect image quality. The Nanaimo lab uses 2 different digital cameras that take excellent images. Follow these links to investigate various specs and options from similar models that you can consider when purchasing your own camera, whatever you make you choose:

Equipment:

• http://www.leica-microsystems.com/ (Leica homepage - international),

http://www.leica-microsystems.com/�


• http://www.leica-microsystems.com/website/products.nsf?opendatabase&language=english

• http://www.qimaging.com/ Qimaging home page

• http://www.qimaging.com/products/cameras/scientific/qicam_fast.php

Camera mounts/adapters that attach the camera to the scope come with/without magnification lenses and are important to consider for best images. The C-mount is probably the most versatile. A good sized computer screen 18 inches or more is rec-ommended for viewing, analyzing and annotating images. A dissecting scope with good quality optics and magnification ability of at least 100-200X is recommended. Objective doublers are useful for capturing detail. Good lighting is also essential. Fibre optics in the form of multiple goose-necks or circles are best for creating either bright and direct/indirect lighting. In order to produce clear detailed photos of high-ly magnified older otoliths, light sources with the ability to vary intensity are needed. This is particularly true for taking photos of older (>30 yrs) redfish otoliths.

Remember to devise a backup storage system for the photo collection. They will take up a lot of storage room and should be kept safe in two locations.

Images of different magnifications of each otolith section are recommended to docu-ment patterns and demonstrate counting rationale. Decide on standard magnifica-tions for capturing whole sections, section halves and close-ups of counting axes or the otolith edge for ease of comparison. Be sure to document this information in a photo log.

Camera & scope/lighting settings:

Software that comes with the camera or separate programs should have the ability to adjust image brightness and contrast of both the live and the snapped images. An auto-contrast option is a time saver. Remember that if an image doesn’t look good/clear down the microscope it is unlikely that you can significantly improve it no matter what tools the software you have offers. Sharp focusing of images to start with are a must. Most software will allow magnification of the live image to check focus.

It is difficult to obtain totally focused images of broken otoliths as their cross-section surfaces are not totally flat. This can be mitigated by using a shallow dish with plas-ticine on the bottom and filled with enough mineral oil or water to cover the otolith. The flat surface of the liquid facilitates focusing of the entire cross-section and elimi-nates “hot spots” produced by multiple fibre optic lighting arms.

Otolith preparation:

Images should be taken using the highest resolution formats possible, e.g. Tiff files. To get the most detail take photos at the highest magnification possible. Images tak-en at low resolutions that are “blown up” to show detail will pixilate and become unclear. Tiff files can be stored as originals, but should be converted to Jpg’s or Pdf’s to reduce their size depending on how and where they are to be used or for sharing via email.

Image format:

An electronic photolog should be maintained to store image information. Commer-cial software is available or devise your own. A database that can be queried is rec-

Metadata:

http://www.leica-microsystems.com/website/products.nsf?opendatabase&language=english�

http://www.leica-microsystems.com/website/products.nsf?opendatabase&language=english�

http://www.qimaging.com/�

http://www.qimaging.com/products/cameras/scientific/qicam_fast.php�


ommended if a high volume of images is expected. Record sample and specimen data (sampling event data, species, age etc.), image/camera/scope settings, photo-grapher, date of photography and reason for taking the photo. Devise a standard meaningful way to name the image files, e.g. redfish1_10yr.tiff (redfish number 1 oto-lith aged 10 years). Choose a naming system that will support a large volume of im-ages over time.

IV: READING

As routine practise, age determination should take place without prior knowledge of fish length to avoid bias. The growth pattern and application of criteria should e the guiding factors, not fish length, deciding age.

It is recommended that the reader use the dorsal side of the redfish otolith section for age determination. The growth pattern on the ventral side is generally less reliable, and should therefore only be used for corroboration.

a) Position angles of light and otolith; reflected versus transmitted light

The angle and intensity of light used to view otoliths depends on the preparation me-thod. Reflected light is best for viewing broken and broken/burnt otoliths. The light burn produced by baking broken otoliths requires more intense transmitted light to effectively illuminate the growth pattern. . However, baking will enhances the sur-face of the otolith cross-section when the otolith is read in transmitted light by keep-ing the surface shadowed. This sentence needs some clarity – I am not understanding the meaning. Thin sections may be read by either transmitted or re-flected light.

b) Axes to count along

The annual growth pattern is often easier to interpret on the otoliths from older or mature fish than those from juveniles. Mature fish otoliths exhibit a pattern "frame of reference" that is not seen on juvenile structures. A reader can see the transition where juvenile annual growth slows down. Because this development is lacking on juvenile otoliths, they are often over-aged as checks are more easily mistaken for an-nuli. Therefore, it is important to keep a picture of a mature adult pattern in mind when interpreting young otoliths or have an older otolith by the scope for quick ref-erence. These changes in pattern deposition are usually easier to interpret on certain areas of the broken or sectioned otolith.

Figure A.2 shows schematically the possible counting axes for broken or sectioned redfish otoliths. Figure A.3 shows the reading axes in accordance to preference for an optimal reading. The dorsal side of otolith sections are favoured for age determina-tion. Axes labelled II (Green lines in Fig A.3) are most preferred and axes I should be avoided (Red lines in Fig. A.3). Checks are most prominent along longest axis.


Figure A.2. Surface of broken or sectioned or otolith (modified after MacLellan, 1997)

Note that axis III near both sides of the sulcus can be confusing due to prominent checks or inconsistencies in pattern. One should read along one of the preferred growth axes that show the most consistent and clearest pattern. Counts should be confirmed by reading as many clear axes as possible. If no axes are clear from nucleus to edge, count in clear areas and use a prominent growth zone to trace from one clear area to another.

Wh en determining the age of older redfish otoliths, some readers find it easier to count from the proximal edge towards the nucleus. In any case, a reader should be able to confirm an age by reversing the direction of a count.

DORSAL(Preferred axes) VENTRAL (Count with caution & beware)

DISTAL (Useful)Longest axis! Do not use!

Figure A.3. Reading axes normally used for redfish age determination (Picture: S. MacLellan)

c) Definition of the nucleus

The nucleus is the central area of the redfish otolith bounded by the first translucent zone (check). It is the central area of the otolith formed during the larval stage. The nucleus in North Atlantic redfish is always opaque.


d) Determination of the first annulus

It is very useful to have otoliths from known-age (e.g., following a strong year-class through its length-modes) juveniles on hand in order to determine the size range and shape of the first year's growth on cross-sections. This will help to establish the loca-tion of the first annulus. Size of the first annual growth zone differs between species, but can also differ between individuals, year-classes and stocks. Shape combined with relative size however, is often characteristic for each species.

Procedures useful in helping to locate and identify the first annulus

1. For broken otoliths, line up the broken otolith halves with the distal surface of one close to the broken surface of the other of known age. Then compare the extent of the first year’s growth (Fig. A.4).

2. Use a micrometer in one of the oculars. Measure the first year's growth as the distance across the nucleus from the dorsal tip to the ventral tip of the annu-lus (see double arrows marking first annulus in Figure A.2). As an example, refer to Nedreaas (1990) who showed the first 5 annuli for S. mentella.

3. Use the same magnification when comparing first annulus measurements of different otoliths. It is important to remember that there will be between-specimen variability so a range rather than particular distance should be de-fined. There will also be differences between stocks/species.

4. Use a convenient mark on a probe or serrations on a forceps as a "measuring stick." By doing this it doesn't matter what magnification is used. The cau-tions given in #3 above also apply here.

Inexperienced readers should get used to finding the first annulus at one magnifica-tion, preferably a lower power, 10-20 x. magnification can be increased once the read-er has counted out to the transition zone where annual growth begins to slow down.

27

31

4

5

8

2

3

8

15


Figure A.4. Otolith collection of known age where the first annulus is known (red arrow). Picture by S. MacLellan

e) Opaque and translucent zones

Together, an opaque zone and a translucent annulus form one year’s growth. Opaque zones found on younger (juvenile) fish otoliths are much larger than those found within the annual growth zones of older otoliths. Checks are often prominent in ju-venile opaque growth zones. Juvenile opaque zones gradually decrease in size to-wards the transition and mature zones, and this should be kept in mind when reading to avoid including checks.

f) Checks

Translucent zone(s) that forms within the opaque (summer) zone represent a slowing of growth. These zones are usually not as prominent as annuli and should not be in-cluded in the age estimate. More than one check per year may form, especially in ju-venile growth zones where they are most prominent.

A check is usually discontinuous, or merges with annuli. In many cases these checks are most prominent on the distal side of the cross-section. They are often not as visi-ble on the proximal side of the boundary (Fig. A.5). Checks usually merge with annu-li at the point where the annulus bends towards the sulcus and continues parallel to proximal edge. On burnt sections both annuli and checks will burn dark.

Not as prominentAre discontinuousMerge with annuli

A

A

Figure A.5. Presence of several checks between two annuli (A). Picture by S. MacLellan

g) Transition zones


A region of change in an otolith growth pattern represented by several annual zones of intermediate size bound by larger juvenile and smaller mature growth zones (Fig. A.6). It is recognized as a region of significant change in the form (e.g., width or clari-ty) of the annual growth zones. During this period of growth, the width of the annuli and opaque zones become significantly reduced. On redfish otoliths, a transition zone is defined as the region of change from more rapid to slower growth. For some fishes this transition zone has been validated as coinciding with the onset of maturation. This still remains to be validated for North Atlantic redfish. The widths of the annuli and the opaque zones become closer in size in the transition zone. This can be a con-fusing area to count in because of the change in annual zone size.

h) Slow growth zones

Annuli and opaque zones become more similar in size within the slow growth zones on older otoliths. The growth is so small that it is rare to see checks. Oxytetracycline (OTC) tagging of Pacific Ocean Sebastes spp. (e.g., Leaman and Nagtegaal 1987, Wal-lace and Tagart 1994) has shown that all the narrow translucent zones in this part of the broken or sectioned otolith should be included and counted as true annuli.

Figure A.6. S. marinus of age 37 years showing the transition zone (light arrows). Pic-ture by S. MacLellan

i) Edge growth

Proper identification of the amount and type of growth on an otolith’s margin is ne-cessary in order to assign the correct age-class. The amount of new (plus) growth on the edge must be related to the time of year the fish was caught and the international-ly accepted and standard January 1st birthday (see Box 1). It is therefore important to have available, during any age reading work, information on the time (day and month) of capture for samples.

Usually, juvenile fish have a proportionately larger amount of growth on their oto-lith's edge than older fish, at any time of the year. Also, juvenile otoliths tend to start showing new growth sooner in the year and may continue to show new edge growth later in the year than what is seen on adult otoliths. Mature fish are redirecting ener-gy towards reproduction for much of the year (development of gonads, spawning and recovering).


Often new growth is hard to see on a broken surface or section's preferred counting axis because it presents a more "compacted" dimension than that of the distal growth axis. Edge growth may be very difficult to interpret on older otoliths at any time of the year as the growth zones are so small. The reader should try to trace the last an-nulus seen from the sulcus area, along the proximal edge to the dorsal tip where the new year's (summer) growth will be most visible. New growth shows ups sooner on the dorsal tip along the distal axis first. In the sulcus area the new year's growth may only appear to be an extra thickening of the last year's annulus.

A problem with the broken/burnt technique is that often, whether or not there is an annulus on the edge, the edge burns dark. Possibly this is because the edges are more directly exposed to the flame, Be aware of this and take into account the time of year and what the unburnt distal surface edges look like as a reference. This should help to interpret broken/burnt otoliths correctly.

In some redfish stocks (e.g., Norwegian and Barents Sea stocks of S. mentella and S. marinus) the new year's growth (plus growth, summer zone) may first appear on the otolith surface in June. In the months January-May it may, however, be difficult to detect the annulus at the edge (separate from the edge itself). Therefore, in the period January 1st - May 31st the edge itself has been counted as being the last year's annu-lus


Box 1. International Birthdate and Age Designation Convention: DFO/Nanaimo

By international convention (Williams and Bedford, 1974), the birthdate of fish has been assigned as Janu-ary 1st, regardless of hatch date. A fish age-reader must assign a fish to its proper age class according to the date it was caught with reference to this birthdate. To do so the reader must judge in which year growth seen on the edge of an otolith was formed. At some times of the year, this decision is relatively straightforward. For example, if an otolith from a fish caught December 31st shows either a full year’s opaque growth or an annulus (translucent material) forming on the margin; the reader would automati-cally attribute this growth to the catch year. Thus, the edge growth would not be counted in the fish’s age. If this same fish were caught one day later, January 1st, it is assigned an age which is one year older. The edge growth is attributed to the previous growth year and therefore is counted. Assigning otolith edge growth from a fish captured from May through August may not be as straightforward and requires knowledge of typical growth tendencies of specific species. See the figure below for further explanation.

These drawings represent otoliths showing growth stages from one January to the next. Growth timing by month may very somewhat between species, stocks, years or locations. The solid black lines represent the translucent annuli (slow growth) which alternate wide white zones represent opaque zones (fast growth). The dashed lines designate incomplete opaque growth. The intermediate growth stage from July through December may show a varying amount of new opaque growth on the otolith margin. The 3rd stage pre-sents two possibilities where an annulus may or may not have started to form late in the year following the deposition of lots of opaque growth. Regardless of the type of growth (opaque or translucent) formed at the otolith margin, the age class advances by one year upon January 1st. The number in front of the brack-ets is AGE CLASS; the number within brackets is annuli seen (partial or complete).

References:

Williams, T . and B.C. Bedford. 1974. The use of otoliths for age determination. p.114-123. T .B. Bagenal [ed.] The ageing of fish. Proceedings of an International Symposium. Unwin Brothers Ltd., Surrey, England. Taken from “Rockfish Age Determination in the Fish Agein g Lab, DFO, Pacific Biological Station, Nanaimo, B.C., Canada”, Shayne MacLellan and “Manual on generalized age determination”. C.A.R.E. 2006 (http://care.psmfc.org)

http://care.psmfc.org/�

Documents

ICES WKADR REPORT 2008 Reports/Expert... · 2013. 2. 4. · ICES WKADR REPORT 2008 3 . 1.3 Workshop structure a nd working procedure Participants are listed in the following Section