This article was downloaded by: [University of Pennsylvania]On: 08 October 2013, At: 13:45Publisher: Taylor & FrancisInforma Ltd Registered in England and Wales Registered Number: 1072954 Registeredoffice: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

New Zealand Journal of Marine andFreshwater ResearchPublication details, including instructions for authors andsubscription information:http://www.tandfonline.com/loi/tnzm20

Age validation and growth of southernblue whiting, Micromesistius australisNorman, in New ZealandS. M. Hanchet a & Y. Uozumi ba National Institute of Water & Atmospheric Research Ltd ,P.O.Box 14–901, Kilbirnie, Wellington, New Zealandb National Research Institute of Far Seas Fisheries , 7–1 Orido 5Chome, Shimizu, Shizuoka‐shi, 424, JapanPublished online: 30 Mar 2010.

To cite this article: S. M. Hanchet & Y. Uozumi (1996) Age validation and growth of southernblue whiting, Micromesistius australis Norman, in New Zealand, New Zealand Journal of Marineand Freshwater Research, 30:1, 57-67, DOI: 10.1080/00288330.1996.9516696

To link to this article: http://dx.doi.org/10.1080/00288330.1996.9516696

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the“Content”) contained in the publications on our platform. However, Taylor & Francis,our agents, and our licensors make no representations or warranties whatsoever as tothe accuracy, completeness, or suitability for any purpose of the Content. Any opinionsand views expressed in this publication are the opinions and views of the authors,and are not the views of or endorsed by Taylor & Francis. The accuracy of the Contentshould not be relied upon and should be independently verified with primary sourcesof information. Taylor and Francis shall not be liable for any losses, actions, claims,proceedings, demands, costs, expenses, damages, and other liabilities whatsoeveror howsoever caused arising directly or indirectly in connection with, in relation to orarising out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Anysubstantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,systematic supply, or distribution in any form to anyone is expressly forbidden. Terms

& Conditions of access and use can be found at http://www.tandfonline.com/page/terms-and-conditions

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

New Zealand Journal of Marine and Freshwater Research, 1996: Vol. 30: 57-670028-8330/96/3001-0057 $2.50/0 © The Royal Society of New Zealand 1996

57

Age validation and growth of southern blue whiting,Micromesistius australis Norman, in New Zealand

S. M. HANCHETNational Institute of Water & AtmosphericResearch LtdP.O.Box 14-901, KilbirnieWellington, New Zealand

Y. UOZUMINational Research Institute of Far Seas Fisheries7-1 Orido 5 Chome, ShimizuShizuoka-shi 424 Japan

Abstract Over 3000 otoliths and 100 000 lengthand sex records were examined of southern bluewhiting caught from commercial fishing groundson the Campbell Island Rise, south of New Zealand,during August and September between 1981 and1989. Ages of juveniles were validated by followingmodes in length-frequency data over a 13-monthperiod between September 1981 and 1982. Agesof adults were validated up to at least 10 years byfollowing strong year classes both from otolith-based age frequency distributions and from length-frequency data from 1981 through to 1989.Independent analysis of the length-frequency datausing MULTIFAN further supported the adultotolith ages. Above age 10 there was less confidencein otolith ages, less agreement between readersand a greater degree of bias between readers. Whenusing the data in catch-at-age models it isrecommended that ages greater than 10 be groupedas a single plus group.

Keywords Micromesistius australis; southernblue whiting; age validation; growth; otoliths;MULTIFAN; length-frequency analysis

M95031Received 16 May 1995; accepted 10 October 1995

INTRODUCTION

Southern blue whiting (Micromesistius australisNorman) is largely confined to subantarctic watersoff the coasts of South America and to the south ofNew Zealand. Its centres of abundance in NewZealand are on the southern Campbell Plateau tothe east of Campbell Islands, Pukaki Rise, andBounty Platform, in depths of 300-600 m (Fig. 1).A fishery developed for this species in the early1970s, and since then annual landings have rangedfrom 3000 to 76 000 tonnes (Hanchet 1993). From1981 to 1989 the fishery focused on an area to thenorth of the Campbell Island Rise (Fig. 1) wherefish gather each August and September to spawn.Annual catches averaged about 15 000 tonnes,increasing in 1989 to about 25 000 tonnes, when itwas New Zealand's fourth largest fishery. Despitethe relatively large size of the fishery little wasknown about the stock size or sustainable yields.During this period otoliths and length-frequencydata were collected from the fishing grounds foruse in catch-at-age models, such as cohort analysisor virtual population analysis (Pope 1972). Severalauthors have recently emphasised the importanceof using reliable (preferably validated) ages whenapplying this technique (e.g., Kimura 1989).

There are several methods of age validation—both direct and indirect (Beamish & McFarlane1983). Direct methods such as captive rearing, tagrecapture studies, and the capture of known agefish are probably not applicable to southern bluewhiting, which are difficult to catch alive and losescales easily on capture. Indirect methods includeanalysis of length-frequency modes, monitoring ofstrong year classes, examination of marginalincrements on hard parts, growth analysis, andcomparison of the different techniques (Beamish& McFarlane 1983).

Previous aging studies of southern blue whitingin New Zealand waters have not successfullyvalidated age estimates (Paul 1992). During the1970/71 cruise of Kaiyo Maru, a maximum age of12 years was recorded using whole otoliths

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

58 New Zealand Journal of Marine and Freshwater Research, 1996, Vol. 30

mmmm PUKAKI RISE

CAMPBELL ISLAND RISE

180°

BOUNTY PLATFORM

Fig. 1 Main distribution area and sampling locations of southern blue whiting.

(Anonymous 1972). Shpak (1978) recorded amaximum age of 14 years from whole otoliths andscales. Van den Broek (1983) prepared and readthin sections of over 500 otoliths and recorded amaximum age of 23 years. Nakaguchi &Shimomura (1989) examined cross sections of over1000 otoliths and recorded a maximum age of 18years. Barrera-Oro & Tomo (1988) examined thincross setions of over 1000 otoliths fromM australisin the south Atlantic around the Falkland Islandsand recorded a maximum age of 23 years. Theyalso published photographs identifying annuli(defined as narrow translucent zones when observedunder reflected light), check rings, and regionswhere annuli could not be clearly discerned.

The objective of the present paper is to validatethe ages of southern blue whiting from New Zealand

waters. This is carried out by following theprogression of modes in juvenile length-frequencydata, and by tracking strong year classes in adultfish using both length-frequency data and annulicounts in otoliths.

MATERIAL AND METHODS

Length-frequency data and otoliths for 1982-86were obtained during cruises of the Japanese MarineResearch Centre's research vessel, Shinkai Maru.Samples were also collected by observers on boardSoviet and Japanese trawlers fishing southern bluewhiting commercially between 1986 and 1989,and by MAF Fisheries staff on board the Sovietcommercial fishing vessel MYS Kuznetsova in 1981(MAF Fisheries unpubl. data). All vessels used

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

Hanchet & Uozumi—Age validation of southern blue whiting 59

60 mm mesh cod-ends, which equates to a 50%selection length for southern blue whiting of about30 cm (Anonymous 1978).

Adult fish were collected during late Augustand September of each year, from the main fishinggrounds north-east of Campbell Island Rise indepths of 400 to 600 m (Fig. 1). Samples of juvenilefish were collected by Shinkai Maru in the samegeneral area throughout much of the year betweenSeptember 1981 and September 1982. Additionalsamples of juvenile otoliths were collected onShinkai Maru in December 1982 and 1984, byobservers in September 1988 and by MAP Fisheriesstaff on board Amaltal Explorer during a researchtrawl survey in October 1989.

Each day up to 300 randomly sampled fishwere measured and sexed, and between 5 and 20pairs of non-randomly sampled otoliths were taken.Sex and fork length (FL, measured to the nearestcm below) were recorded for each fish. No otolithswere collected from the commercial fishery in1987.

Length-frequency dataLength-frequency data were collected by scientistson board Shinkai Maru from 1982 to 1986 and byobservers from 1986 to 1989. Length-frequencydata for the year of overlap of the Shinkai andobserver data sets were near-identical (Hanchet1991). The data sets were therefore combined foranalysis.

Length-frequency data were analysed to deriveestimates of the growth curve parameters andlengths at age independent of the otolith ages, vonBertalanffy growth curves were fitted to the maleand female length-frequency distributions separ-ately using the MULTIFAN model Version 32(Fournier et al. 1990). This model simultaneouslyanalyses multiple sets of length-frequency samples,using a maximum-likelihood method to estimatethe proportions of fish in each age class, and thevon Bertalanffy growth parameters Lx, K, and to-As southern blue whiting are known to spawnduring September of each year (Hanchet 1993),the fish were assigned a birthdate of 1 August, andthis was used as the birth date to calculate ?o fromthe age of the first year class in the first sample, towas estimated using the equation:

tc, ="K

where t\ is the estimated age (in years from thebirth date) of the first age class and m\ is the mean

length of that age class, as estimated byMULTIFAN.

There is a danger that parameter estimates madeby MULTIFAN will represent only a localminimum. To overcome this problem, we startedby making a systematic search, using all combin-ations from a matrix of lvalues (0.20, 0.25, 0.30,0.35, 0.40, 0.45) and plausible age classes (usually10 to 16). For the analysis of each sex the procedurewas to start with the simplest model and then togradually increase the complexity by introducing(i) variable standard deviation and (ii) constraintson mean lengths of ages 1 and 2. At each stage theparameters were estimated, with K always beingthe last parameter to be estimated. Log-likelihoodratio tests were carried out to determine whetherthe inclusion of extra parameters significantlyimproved the model. Following Fournier et al.(1990), we used a significance level of 0.10 fortesting whether there is any gain in introducing anadditional age class in the length-frequencyanalyses; all other tests were carried out with asignificance level of 0.05.

OtolithsOne of each pair of otoliths was burnt over analcohol flame, and embedded in a resin block whichwas then cut through the core using a low-speeddiamond saw. The cut surface of the section of theotolith was viewed under reflected light.

Age estimates were determined by countingthe number of annuli on the cut surface of theotolith. The annulus is defined as the translucentzone, or the zone of slower growth, that appearedas a dark zone under reflected light. Zone countswere usually made from at least two different areasof the otolith:

(1) from the core out to the dorsal edge, and(2) from the core out to the proximal edge (areas I,

II, and IV in Fig. 2), following Chilton &Beamish (1982).

Most otoliths collected during August andSeptember had wide opaque zones at the edge ofthe otolith, although occasionally a narrowtranslucent zone had begun to form. Because thebirthdate was assigned 1 August one was added tothe age estimate of all fish caught during Augustand September which had an opaque zone at theedge of the otolith.

Otolith appearance and structure was verysimilar to that reported by Barrera-Oro & Tomo(1988), and so interpretation of the annuli followed

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

60 New Zealand Journal of Marine and Freshwater Research, 1996, Vol. 30

Distal edge

Checks and annuliCore

First annulus

Fig. 2 Generalised cross sectionof an otolith showing selectedcounting areas (after Chilton &Beamish 1982).

m iv

Proximal edge

their methodology. The first annulus was oftenabsent or poorly defined so back-calculation wasapplied to predict the likely position of the firstand second annuli. In back-calculation proceduresthe diameter of the zone was measured in thetransverse plane from the ventral to the dorsaloutside edge. Care was also taken to correctlydistinguish annuli from the check rings whichoccurred occasionally between the core and thefirst, second, and third annuli. In general, thesecheck rings were identified as being narrower thanthe annuli and usually could not be traced aroundthe entire core.

A readability category ranging from 1 to 5 wasrecorded for each otolith:

1—Clear and unambiguous;2—Little doubt;3—Reasonably confident but some doubt;4—Considerable uncertainty;5—No confidence in age estimate.

A total of 3218 otoliths were read. The sex andlength of the fish that the otolith came from wasnot known to the reader. No attempt was made torandomise the samples by year because the agereaders did not know that strong year classeswere present. A random sample of about 40otoliths from the years 1983 to 1989 was read by asecond reader to examine "between-reader"variability.

Growth curves were fitted to the otolith agesusing the non-linear multivariate secant parameterestimation procedure (SAS Institute 1988).

Differences in growth between the sexes wereexamined using the log-likelihood ratio test.

Age data were scaled up to catch-at-age becauseotoliths were not always collected in a randommanner. As the majority of the catch was taken inAugust and September (Hanchet 1993), catch-at-age estimates for each year were derived bycombining the age-length key for that year withthe weighted length-frequency for that year. Thesexes were treated separately but combined toproduce the final age-frequency distributions. Theweighted length-frequency data were obtained inthe following way. First, the data were stratifiedinto weekly time periods. Then the length-frequencysamples taken aboard the vessels were scaled up tothe total weight of the sampled catch. These werescaled up to the weight of the catch taken by thewhole fishery for that week. Results from eachweek were then summed to give the weighted length-frequency distribution of the total catch for theyear.

RESULTS

Juvenile length-frequency distribution andotolithsThe monthly progression of modes in length-frequency data can be seen in Fig. 3. The mode at12 cm in April 1982 can be followed through to asmall mode at 14—17 cm (represented by only 20fish) in June 1982. A mode at about 22 cm inSeptember 1981 can be followed through to 29 cm

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

Hanchet & Uozumi—Age validation of southern blue whiting 61

Fig. 3 Length-frequency distri-bution of small fish (sexescombined) from southernCampbell Plateau, September1981-1982. Sample size is in themiddle. The size distributionshave been arbitrarily cut off at 40cm.

40

O 30-

? 20CD

5 10--'81Sep Oct Nov

n=365

Dec

n=204

'82Jan Feb

-V

Mar

n=204

Apr

0+

May

>--

n=1853

Jun

1-1409

Jul

n=4945

Aug

n 2588

Sep

in July 1982, and a mode at 30 cm in September1981 can be followed through to 33/34 cm inSeptember 1982. On the basis of the length-frequency data, and the perceived early growth ofthe fish, the smallest three modes in Fig. 3 wereassigned the ages 0+, 1+, and 2+.

Otoliths were not collected from most of thesamples in Fig. 3. However, samples collected inother years had modes at very similar lengths tothose in 1981-82 (Fig. 4). There was also goodcorrespondence between annuli counts on theseotoliths and the ages determined from the modal

ApriM 982 (n = 138) October 1989 (n = '

40 •

30

20

10

0

I I I I I I I I I I I I I I I I I I I I I I

September 1988 (n = 145)

• 0• 1

03ID 4

I I I I I T I I I I I I I I I ITTT I I I10 15 20 25 30

Fork length

|16-14-12-10-8-6-4-2-

n nl In •ft n

12-

10

8-

6-

4-

2-

I I I I I I n I I I T IT I T I I I I I I I

December 1982 and 1984 (n = 50)

I I I

I I I I I II I I I I I •10 15 20

Fork length

Fig. 4 Correspondence between hyaline ringgroups and fork length for southern blue whiting <37 cm FL collected from the southern CampbellPlateau.

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

62 New Zealand Journal of Marine and Freshwater Research, 1996, Vol. 30

Males Females

I

20 30 40 50 60

Fig. 5 Length-frequency histograms of southern blue whiting collected from the Campbell Island commercialfishery each September from 1981 to 1989, showing the best fit of MULTIFAN model. Curves are shown for eachage class separately (normal curves) and for the sum of all age classes (uppermost curve), n, sample size. Left, male;right, female.

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

Hanchet & Uozumi—Age validation of southern blue whiting 63

size classes for small southern blue whiting. Fishof the first mode in April 1982 had no annulivisible, those in the second mode had 1 annulusvisible, whereas those in the third mode had 2annuli visible. Although checks often occurredbetween the first and second annuli these couldusually be identified because they were narrowerand less distinct than the annuli. Length modesremained reasonably discrete up to at least age 2.

Adult length-frequency distributionThe results of the best MULTIFAN fits to the dataare shown in Table 1. For males, constant standarddeviation with 13 age classes gave the best model.For females, variable standard deviation with 13age classes gave the best model. The raw data withthe normal curves overlain are shown in Fig. 5.

Table 1 von Bertalanffy parameter estimates (withstandard errors) for the best-fit MULTIFAN models.

Standard deviationMales

ConstantFemalesVariable

Age classesAT (year1)

toAverage SDRatio SD

130.36 (0.17)

46.70 (0.19)

-0.01 (0.17)1.64 (0.03)1.00

130.40 (0.11)

49.40 (0.17)

-0.19 (0.14)2.28 (0.15)0.77 (0.02)

The raw data show the annual progression of adominant mode from a length of 30 cm in 1981through to 46 cm for males, and 49 cm for females,in 1989. More recently modes can be tracked insmaller males from 1985 and 1988 through to 1989,and in smaller females from 1986 to 1989.

In general, the normal curves from the bestMULTIFAN fit show good correspondence withthe raw length-frequency data, and track thedominant mode reasonably well. The MULTIFANgrowth curves for either sex are compared in Fig. 6with the modal lengths at age of juveniles (sexescombined) caught in April 1982 (from Fig. 3). Theearly part of the female MULTIFAN growth curveclosely fits the juvenile ages, reaching 29 cm FL atage 2, whereas the early part of the MULTIFANmale growth curve predicts a slower growth rate,reaching only 24 cm FL at age 2.

Adult otolithsThere were significant relationships between logotolith width and fish length (P < 0.05), and logotolith thickness and fish length (P < 0.05) (Table2). The back-calculated size of the first annuluswas found to be in good agreement with the rangein otolith widths measured from fish of the smallestsize class collected from September, when theannulus appears to be deposited. It also showedlittle overlap in distribution with the back-calculatedsize of the second annulus.

Table 2 Linear regressions of log otolith width and thickness on log forklength, n, number offish.

Sex n Intercept Slope R2

Otolith

Otolith

width

thickness

MF

MF

14851436

14841437

1.1571.166

-0.059-0.084

0.00850.0085

0.00440.0043

0.880.88

0.740.72

<0.05<0.05

<0.05<0.05

Table 3

RC 1

Percentage

2 3

offish at

4

each age

5

in each

6

readability category,

Age

7 8 9

RC.

10 1 1 12 13 14+ Total

12345

Total

7525

4

44331355

128

25501762

390

204723

82

362

205220

71

380

10492911

1

297

84434131

171

92737189

158

62936235

184

2244026

8

101

228263410

61

37332010

30

328381813

61

22745179

406

3831101780351118

2733

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

64 New Zealand Journal of Marine and Freshwater Research, 1996, Vol. 30

60

50

40

30

20

10

0

f

P

M

50

40

30

20

.B.Jl-0-o o_~-Q-o~----e—-<,-• F

M

Fig. 6 MULTIFAN growth curves for male (M) andfemale (F) southern blue whiting. Also plotted are themodal lengths at age for 0+, 1 +, and 2+ fish caught inApril 1982 (from Fig. 3).

Fig. 7 Growth curves of males (M) and females (F)using all otoliths collected from August and September1983 to 1989. +, observed mean male lengths at age; o,observed mean female lengths at age.

Growth rates were found to be significantlydifferent between sexes (%2 = 997', d.f. = 3,P< 0.001), and so were analysed and plottedseparately (Fig. 7). Growth of the fish is relativelyfast over the first 5 years, slows down over the next5 years and virtually ceases after age 10. Themaximum ages recorded were 24 and 25 for malesand females, respectively, but ages greater than 2 ]were uncommon. The maximum observed lengthsof fish in the population are 54 and 59 cm formales and females, respectively (MAF Fisheriesunpubl. data), although few male and female fishare caught > 49 and 53 cm FL, respectively (seeFig. 5).

The reader was reasonably confident of 83% ofthe age estimates and had no confidence in only4% (Table 3). There was less confidence in ageestimates from older fish, with a marked increasein uncertainty after age 7.

There was a good level of agreement betweenreaders in the younger fish (Table 4), althoughthere was a tendency for Reader 2 to underestimateages compared to Reader 1, particularly in fisholder than age 10. Over 88% of ages agreed to ± 1year in readability categories 1-5, increasing to95% agreement to ± 1 year for categories 1—3.

The age distribution of the commercial catch isshown for males and females in Fig. 8. From 1983

to 1988 the catch was numerically dominated bytwo strong year classes spawned in 1979 and 1980.These year classes can be followed from ages 3and 4 in 1983 through to ages 9 and 10 in 1989.Because of their dominance in the catch it is difficultto track other strong year classes in the older fish,but these possibly occurred in 1965, 1971, and1975. The 1986 year class also appears to bemoderately strong.

DISCUSSION

Validation

Beamish & McFarlane (1983) outlined severalindirect methods of age validation which includedanalysis of length-frequency modes, monitoring ofstrong year classes, growth analysis, andcomparison of the different techniques. Probablythe most convincing of these is the ability to followstrong year classes either through otolith annulicounts or length-frequency data.

In the present study ages of juveniles werevalidated by following modes in length-frequencydata over a 13-month period between September1981 and 1982. Otolith ages in the adult fish areconsidered to be validated up to at least 10 yearsby the ability to follow strong year classes both

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

Hanchet & Uozumi—Age validation of southern blue whiting 65

o

ICO

1 =

'80

10-

5-

0

J.79 1983H N = 413l l 7,5 71 I

i I I I I I I i I I i I I

15-

10-

5-

0

1984N = 387

-II- ii I I I I I I I I I I I I I I I I

15-

10-

5-

0

1985N = 385

1 1 I I I I I I I I 1 I I I 1 I 1 I 1 I

o

ber

t2

1 5 "

10-

5 -

0

HI I I I I I I I I I 1 1 | | 1

1986N = 466

i i i i

15-

10-

5-

0

1988N = 548

.111.1.15-

10-

5 -

0

i I I I I I I I I I I I I I I I I I I I

'86

ll.l.h1989

499

I I T T I f i i • • i i i i i i i i i i

2 4 6 8 10 12 14 16 18 20+Age (years)

Fig. 8 Age distribution of the catch between 1983 and1989 (sexes combined). Pointers indicate possible strongyear classes and the year in which they were spawned.

from otolith-based age-frequency distributions andfrom length-frequency data from 1981 through to1989. Growth parameters and mean lengths at agederived independently from the analysis of length-frequency data using MULTIFAN are very similarto those derived from the otolith ages in this study(Table 5, Fig. 9).

At first sight the presence of two strong yearclasses observed in the age distribution data

between 1983 and 1989 appears to conflict withthe single mode visible in the length-frequencydistribution of the commercial catch over this sametime period (Fig. 5). However, there is evidencefrom the juvenile length-frequency data that therewere two strong year classes in early 1982 (Fig. 3).The two modes at 27-28 cm and 32-33 cm in April1982 occurred in most tows on the CampbellPlateau and Pukaki Rise during a random trawlsurvey of the area (van den Broek et al. 1984). Thesize group of smaller fish did not recruit into thespawning fishery in 1982, because most fish do notmature until they are 3 years old (Hanchet 1993),and hence was not seen in the length-frequencydistribution of the commercial fishery (see Fig. 5).However, by 1983 both year classes are at leastpartially recruited to the spawning fishery andbecause of the range of growth rates within each ofthe two year classes the length-frequency dataappears as a single mode.

Length-frequency and year class strength datasuccessfully validate the ages of southern bluewhiting at least up to an age of 10 years. Aboveage 10 there was less confidence in otolith ages,there was less agreement between readers, andalso a greater degree of bias between readers. Forthe purposes of using the data in catch-at-agemodels it is recommended that ages greater than10 be grouped as a single plus group.

Comparison of growth between studiesGrowth parameters from both the MULTIFAN andthe otolith data sets in the present study are in closeagreement with those of van den Broek (1983),having relatively high K values, and differ fromthe earlier Japanese studies (Anonymous 1972;Nakaguchi & Shimomura 1989) which had highervalues of L^, but lower values of K (Table 5).However, all curves show very similar trajectories,with rapid growth over the first 5 years, a slowingdown over the next 5 years, and almost a cessationof growth after age 10 (Fig. 9A, 9B). In commonwith the other studies, females were found to growsignificantly faster than males and reached a largeroverall length.

Maximum ages from otoliths in the presentstudy were similar to those of van den Broek whoused thin sections, but considerably higher than inthe other studies. The earlier studies of Anonymous(1972), and Shpak (1978) used whole otoliths andscales, both of which are known to underestimateages (Beamish & McFarlane 1987). Estimates oflongevity from MULTIFAN are likely to be

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

66 New Zealand Journal of Marine and Freshwater Research, 1996, Vol. 30

5OH

o 30

S20-

0{

Age (years) Age (years)

Fig. 9 von Bertalanffy growth curves for New Zealand southern blue whiting. 1, Anon 1972; 2, van den Broek1983; 3, Nakaguchi & Shimomura 1989; 4, Present study—otoliths; 5, Present study—MULTIFAN. A, Males; B,Females.

Table 4 Frequency of agreement of all otolith ages between two readers by age. Rl-R2, differences in agesbetween readers 1 and 2 (i.e., 0 implies exact ageement between the two readers).

R1-R2

_2-1

0+1+2+3

>+3

2

16

3

441

4

2264

5

134

8

6

2155

Age

7

2851

recorded by

8

11842

9

242

reader 1

10

33

11

13

1

12

13

121

13

1111

14+

163

141045

Total

325

150501665

Table 5 Age and growth studies of southern blue whiting from New Zealand waters. Method: W, whole otoliths;TS, thin sections; XS, cut surface of block.

Source

Anon (1972)Shpak(1978)

van den Broek(1983)

Nakaguchi &Shimomura (1989)

Present study

Method

Otoliths (W)Scales and

otoliths (W)Otoliths (TS)

Otoliths (XS)

Otoliths (XS)

MULTIFAN

48.6—

45.5

50.2

46.5

46.7

Males

K

0.31—

0.38

0.29

0.39

0.36

to

—0.02—

-0.48

-0.83

-0.68

-0.01

Max.age

1014

22

16

24

13

53.1—

48.8

52.5

50.1

49.4

Females

K

0.26_

0.34

0.27

0.35

0.40

'0

-0.02—

-0.54

-0.89

-0.71

-0.19

Max.age

1214

23

18

25

13

Samplesize

_

554

1116

2787

-

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3

Hanchet & Uozumi—Age validation of southern blue whiting 67

minimum estimates due to the lumping of the olderyear classes into a single mode. The differencebetween this study and that of Nakaguchi &Shimomura (1989) lay chiefly in the interpretationof the zones of the older fish. In the present study itwas assumed that all translucent zones after age 4were formed annually, whereas in the earlier studyit was assumed that check rings still occurred afterthis age (N. Nakaguchi, pers. comm., 1990). Giventhe validation of the annual nature of zone formationup to at least age 10, it seems that the formerassumption is more likely to be correct.

ACKNOWLEDGMENTS

We are grateful to many of the staff at the NationalResearch Institute of Far Seas Fisheries for help duringthe study. In particular, we thank M. Sugiyama forassistance with otolith preparation. The first authoracknowledges the Science and Technology Agency ofJapan for financial support during this study. Thanksalso to K. Lister for reading a subsample of the otolithsand to G. MacKay and J. Ingerson for drafting severalfigures. J. Kalish, S. MacLellan, and R. Stanley madeconstructive comments on earlier drafts of this paper.

REFERENCES

Anonymous 1972: Report of an investigation by theKaiyo Maru in New Zealand waters in 1970-71.Fishery Agency of Japan, 292 p. and 290 p. oftables and figures (in Japanese). (Translation104 held in NIWA Greta Point library,Wellington).

Anonymous 1978: Report of survey cruise of the KaiyoMaru in New Zealand waters, 1977-78. FisheryAgency of Japan, 150 p (in Japanese).(Translation 97 held in NIWA Greta Pointlibrary, Wellington).

Barrera-Oro, E. R.; Tomo, A. P. 1988: New informationof age and growth in length of Micromesistiusaustralis Norman, 1937 (Pisces, Gadidae), inthe South-West Atlantic. Polar biology 8: 341-351.

Beamish, R. J.; McFarlane, G. A. 1983: The forgottenrequirement for age validation in fisheriesbiology. Transactions of the American FisheriesSociety 112:735-743.

Beamish, R. J.; McFarlane, G. A. 1987: Current trendsin age determination methodology. In:Summerfelt, R. C.; Hall, G. E. ed. Age andgrowth of fish. Iowa State University Press,Ames, pp. 15-42.

Chilton, D. E.; Beamish, R. J. 1982: Age determinationmethods for fishes studied by the groundfishprogram at Pacific Biological Station. Canadianspecial publication of fisheries and aquaticsciences 60: 102 p.

Fournier, D. A.; Sibert, J. R.; Majkowski, J.; Hampton,J. 1990: MULTIFAN: a likelihood-based methodfor estimating growth parameters and agecomposition from multiple length-frequency datasets illustrated using data for southern bluefintuna (Thunnus maccoyii). Canadian journal offisheries and aquatic sciences 47: 301-317.

Hanchet, S. M. 1991: Southern blue whiting fisheryassessment for the 1991-92 fishing year. NewZealand Ministry of Agriculture and Fisheries,Fisheries assessment research document 91/7.48 p.

Hanchet, S. M. 1993: Southern blue whiting fisheryassessment for the 1993-94 fishing year. NewZealand Ministry of Agriculture and Fisheries,Fisheries assessment research document 93/17.56 p.

Kimura, D. K. 1989: Variability in estimating catch-in-numbers-at-age and its impact on cohort analysis.In: Beamish, R. J.; McFarlane, G. A. ed. Effectsof ocean variability on recruitment and anevaluation of parameters used in stock assess-ment models. Canadian special publication offisheries and aquatic sciences 108. 57-66

Nakaguchi, N.; Shimomura, T. 1989: Age and growthof southern blue whiting (Micromesistiusaustralis pallidus) in New Zealand. UnpublishedBSc thesis, Tokai University, Shimizu, Japan(in Japanese).

Paul, L. J. 1992: Age and growth of New Zealandmarine fishes, 1921-90: a review andbibliography. Australian journal of marine andfreshwater research 43: 879-912.

Pope, J. G. 1972: An investigation into the accuracy ofvirtual population analysis using cohort analysis.International Commission of the NorthwestAtlantic fisheries research bulletin 9: 65-74.

SAS Institute 1988: SAS/STAT User's Guide, Release6.03 Edition. Cary, NC: SAS Institute Inc., 1988.1028 p.

Shpak, V. M. 1978: The results of biologicalinvestigations of the southern putassu Micro-mesistius australis (Norman, 1937) on the NewZealand plateau and perspectives of its fishery.Unpublished TINRO manuscript (in Russian).

van den Broek, W. L. F. 1983: Ageing deepwater species:report of a visit to the UK, Sep-Nov 1982. NewZealand Ministry of Agriculture and Fisheries,unpublished Fisheries Research Division internalreport, 20 p.

van den Broek, W. L. F.; Tokusa, K.; Kono, H. 1984: Asurvey of demersal fish stocks in waters south ofNew Zealand, March-May 1982. New ZealandMinistry of Agriculture and Fisheries, FisheriesResearch Division occasional publication 32.28 p.

Dow

nloa

ded

by [

Uni

vers

ity o

f Pe

nnsy

lvan

ia]

at 1

3:45

08

Oct

ober

201

3