Geographic variation and juvenile migration in Atlantic redfish inferred from otolith microchemistry

CSIRO PUBLISHING

www.publish.csiro.au/journals/mfr Marine and Freshwater Research, 2005, 56, 677–691

Geographic variation and juvenile migration in Atlantic redfishinferred from otolith microchemistry

Christoph StranskyA,D, C.-Dieter Garbe-SchönbergB and Detlef GüntherC

AFederal Research Centre for Fisheries, Institute for Sea Fisheries, Palmaille 9,22767 Hamburg, Germany.

BUniversity of Kiel, Institute of Geosciences, Department of Geology, Olshausenstr. 40,24118 Kiel, Germany.

CLaboratory of Inorganic Chemistry, ETH Hönggerberg, HCI, 8093 Zürich, Switzerland.DCorresponding author. Email: [email protected]

Abstract. Golden redfish (Sebastes marinus) and deep-sea redfish (S. mentella) are heavily exploited fish resourcesin the North Atlantic, but their stock delineation remains unresolved. The present study examined the use ofotolith microchemistry as a stock separation tool for redfish. By determining minor and trace elements in redfishotoliths from various areas in the North Atlantic, geographic and temporal variations in otolith microchemistry wereinvestigated. Relatively high temporal stability in otolith elemental composition was found for juvenile redfish fromnursery areas off East Greenland. Elemental concentrations in the nucleus, juvenile and marginal otolith zonesdiffered significantly between areas and showed consistent longitudinal trends for some elements. Multivariateanalysis of element constituents by area, however, revealed poor geographic separation (<50% cross-validatedclassification) for both species. Otoliths of juvenile S. mentella from demersal nursery grounds and adjacent pelagichabitats showed no significant differences between areas. The results are in accordance with the recently observedlow morphometric and genetic heterogeneity of redfish across the North Atlantic. It is, however, unclear if the lackof clear spatial separation by natal signatures is due to a common origin of the investigated fish or due to limitedvariability of the chemical composition of their ambient environment.

Extra keywords: elemental signatures, LA-ICP-MS, North Atlantic, Sebastes marinus, Sebastes mentella.

Introduction

Golden redfish (Sebastes marinus) and deep-sea redfish(S. mentella) are widely distributed in the North Atlanticand provide important fishery resources. Sebastes marinusis found on the continental shelves off eastern Canada,Greenland, Iceland, the Faroe Islands, Norwegian waters, theBarents Sea and Svalbard, mainly in depths between 100 and300 m (Whitehead et al. 1986). Sebastes mentella is gen-erally distributed deeper than S. marinus and also inhabitsthe pelagic zone of the Labrador and Irminger Sea downto 1000 m (Magnússon and Magnússon 1995). For fisheriesmanagement purposes, several stock units were establishedfor redfish by the International Council for the Explorationof the Sea (ICES) and the Northwest Atlantic Fisheries Orga-nization (NAFO). Sebastes marinus is currently divided intotwo ICES management units, one comprising the occurrencesoff Greenland, Iceland and the Faroe Islands, and the otherone along the Norwegian coast, in the Barents Sea and offSvalbard (ICES 1998). The same management units apply fordemersal S. mentella, whereas pelagic S. mentella is currentlyassessed by ICES and managed by the Northeast Atlantic

Fisheries Commission (NEAFC). In the North-west Atlantic,demersal redfish, not separated by species, are managedwithin NAFO in the offshore areas east of Newfoundlandand Labrador, on the Flemish Cap, off West Greenland, NovaScotia, in the Gulf of Maine and on Georges Bank.

Due to the highly migratory nature of S. mentella, theintraspecific variation and the elucidation of the complexstock structure has gained particular interest, especially withrespect to fisheries management of this valuable resource(ICES 1998). Morphometric studies have attempted to iden-tify sub-units and stocks, reporting large-scale geographicdifferences (Saborido-Rey 1994), but also sub-structure ofS. mentella in the North-east Atlantic (Saborido-Rey andNedreaas 2000). Biochemical and genetic studies showedlow intraspecific variation (Nedreaas et al. 1994), while amore extensive study indicated differences between pelagiccomponents of S. mentella (Johansen et al. 2000). Recentmicrosatellite DNA analyses by Roques et al. (2002) pointto a ‘pan-oceanic’ population, with high within-populationhomogeneity, spanning from the Grand Banks to theFaroe Islands. Recently reported parasitological analyses by

© CSIRO 2005 10.1071/MF04153 1323-1650/05/050677

678 Marine and Freshwater Research C. Stransky et al.

Marcogliese et al. (2003), however, suggested a separation ofS. mentella on the Flemish Cap, in the Gulf of St. Lawrenceand the Laurentian Channel.

Another natural tag that has been widely used in stock sep-aration research during recent years is the elemental composi-tion of calcified structures in the fish body, such as scales andotoliths (see reviews by Campana 1999 and Thresher 1999).Although not always straightforward, the elemental composi-tion of the otoliths reflects that of the ambient seawater duringa certain period of the fish’s life (Elsdon and Gillanders 2003).Because otoliths grow continuously, mainly due to the depo-sition of calcium carbonate and a variety of minor and traceelements, the environmental history of the fish under inves-tigation can be tracked using elemental and stable isotopesignatures. From a variety of instrumentats available for theanalysis of elemental contents in fish otoliths (see Campanaet al. 1997), particle-induced X-ray emission (PIXE) andeven more frequently used, inductively coupled plasma massspectrometry (ICP-MS) have prevailed over more traditionalelectron microprobe techniques (Campana 1999).

By coupling a laser ablation system with ICP-MS(LA-ICP-MS), analyses at varying resolution can be car-ried out within specific regions of the otolith, e.g. from thenucleus (core) to trace the origin of the fish, or the marginalgrowth zones to extract information on the recently inhab-ited environment. The resulting elemental fingerprints canbe compared between geographic areas or habitats across theotolith growth layers to investigate stock origin, mixing andmigration (Campana 1999). Most recent investigations usingthese techniques for the separation of marine fish populationshave focused on coastal or estuarine areas on different spa-tial scales (e.g. Milton and Chenery 2001; Gillanders 2002).Only a few LA-ICP-MS studies were carried out in offshoreor oceanic areas to facilitate stock discrimination of migra-tory species, such as bluefin tuna (Rooker et al. 2003) anddeep-water fish (Gordon et al. 2001).

The present study presents the first attempt to investi-gate otolith microchemistry of Atlantic redfish across thedistributional range of two species with high commercialinterest and unresolved stock structure: S. marinus andS. mentella. The primary aim was to test the use of elemen-tal fingerprints as a stock discrimination tool by analysingthe minor and trace element composition in the nucleusregion of redfish otoliths, giving information on the indi-vidual natal origin. Sebastes marinus otoliths were alsoanalysed within the marginal zones, and otoliths of the highlymigratory S. mentella were sampled within the juvenile andmarginal zones to investigate the environmental history ofthese species. Because the temporal stability of elementalsignatures is an important prerequisite for the validity of thederived stock patterns (Gillanders 2002), otoliths of juvenileS. marinus and S. mentella were sampled in an importantnursery area off East Greenland (Magnússon et al. 1988)over a period of five years and were analysed for inter-annual

variation in element concentrations. By following strongcohorts in length–frequency spectra, evidence for the migra-tion of juvenile S. mentella from the East Greenland shelfinto the pelagic habitat in the Irminger Sea during 1998–1999 was recently found (Stransky 2000). Otoliths collectedin these areas during that period were analysed for elementalcomposition in the nucleus and marginal regions to investi-gate the possibility of tracking redfish migration routes usingotolith microchemistry.

Material and methods

Sample collection

The otolith samples were divided into three sets and analysed for theirmicrochemistry (Table 1). The location of the sampling stations isgiven in Fig. 1. The first sampling set consists of otoliths of juvenileS. marinus and S. mentella (20–25 cm), caught on the Heimland Ridgeoff East Greenland (64◦N 36◦W) over a period of five years (1998–2002) on an annual groundfish survey in autumn. The second set ofotoliths was collected from adult S. marinus and S. mentella with totallengths of >30 cm, taken in various locations across the North Atlantic.Because evidence for an extensive migration of juvenile S. mentella(25–30 cm total length) from East Greenland into the Irminger Sea wasfound during 1998/1999 (Stransky 2000), the available material fromfish collected in these areas during that period was analysed as third setof otoliths. When only fork lengths were available, total lengths werecalculated on the basis of conversion factors for the respective species inthe respective area (F. Saborido-Rey, unpublished data). The Icelandicsamples were taken on a commercial vessel, while all other samples werecollected on research vessels. Bottom trawls were the predominantly-used fishing gear, with the exception of S. mentella in the Irminger Seathat were caught with pelagic Gloria-type trawls. Sample sizes of at least10 otoliths were targeted but not always achieved due to time and budgetconstraints. All redfish otoliths were removed from the fish shortly aftercapture, rinsed in tap water and stored dry in paper envelopes.

LA-ICP-MS analysis

Preceding microchemical analysis, all otoliths were embedded intotransparent polyester resin and thin-sectioned transversely through thenucleus (core) region with a diamond-blade saw to approximately1.0 mm thickness. The thin sections were mounted onto circular glassmicroscope slides, rinsed with Millipore™ filtered Milli-Q water(≥18 M� cm−1) to minimise surface contamination, dried in a laminarflow hood at room temperature and stored in clean sealable plastic bags.The left-sided otolith of an adult S. marinus collected off East Green-land (65◦29′N 32◦31′W, October 2001) was prepared in the same way,but not mounted on a slide. This otolith was used as a solid in-housestandard (Reference Otolith) for the monitoring of long-term repro-ducibility. The right-sided otolith of the same individual was preparedfor solution analysis by grinding the otolith to a core containing thenucleus and partly adjacent growth zones. The concentrations of Li,Cu and Rb obtained for this sample were similar to the median valuesobtained from nucleus LA-ICP-MS, whereas Sr and Ba in the solution-based analysis were markedly higher than in the nucleus ablations. Thisdifference can be in part explained by generally elevated Sr and Ba con-centrations towards the outer regions of the otoliths and the fact that themicro-preparation of the Reference Otolith included also non-nucleus,marginal areas. The accuracy of the solution-based ICP-MS results weretraced back to NIES CRM No. 22, Fish Otolith (Yoshinaga et al. 2000).Our results confirmed the certified values for Zn (0.6 µg g−1; CRM:0.5 µg g−1), Sr (2460 µg g−1; CRM: 2360 µg g−1) and Ba (3.0 µg g−1;CRM: 2.9 µg g−1).

Geographic variation in redfish otolith microchemistry Marine and Freshwater Research 679

Table 1. Overview of otolith samples analysed for microchemistry

Species Area Area code Sampling period Fish total No. oflength (cm) otholiths

S. marinus East Greenland (Heimland Ridge)A Oct 1998–Oct 2001 21–25 18S. mentella East Greenland (Heimland Ridge)A Oct 1998–Oct 2002 20–24 27S. marinus Flemish Cap FC Jul 2001 31–68 10S. marinus West Greenland WG Oct 2000 31–58 10S. marinus East Greenland EG Sep–Oct 2000 31–65 10S. marinus Iceland IS Feb–Mar 2001 31–45 10S. marinus Barents Sea BS Jul–Aug 2000 31–40 10S. mentella Flemish Cap FC Jul 2001 31–43 5S. mentella Davis Strait DS Sep–Oct 2001 36–46 5S. mentella East Greenland EG Oct–Nov 2001 31–41 10S. mentella Irminger Sea IR Jun–Jul 1999 32–47 10S. mentella Iceland IS Oct 2000–Apr 2001 31–50 10S. mentella Faroe Islands FA Sep–Oct 1999 33–49 10S. mentella Barents Sea BS Jul–Aug 2000 32–39 5S. mentella East GreenlandB Oct 1998 25–30 13S. mentella Irminger SeaB Jun–Jul 1999 26–30 10

ASamples of juvenile fish used for the investigation of temporal stability, Bsamples of juvenile fish used for the investigation ofmigration from East Greenland into the Irminger Sea.

The sampling slides, containing 9–20 otolith sections, were analysedfor trace elements using LA-ICP-MS with an Agilent 7500s (Agi-lent Technologies, Tokyo) and a Elan 6100 DRC plus (Perkin Elmer,Norwalk, CT). A 193 nm ArF Geolas M and Q (Microlas, Göttingen,Germany) Excimer laser ablation system was coupled with the ICP-MS instruments to ablate material from the otolith section surface. Thelasers were operated at a wavelength of 193 nm with a pulse rate of10 Hz and an energy output of 135 mJ (Günther and Heinrich 1999).The sample craters produced had a diameter of approximately 31 µm.The ablation cell was flushed with a 1.2 L min−1 He gas flow, transport-ing the ablated material to the ICP. The He carrier gas was mixed witha 0.8 L min−1 Ar gas flow 30 cm in front of the ICP. Signal intensitieswere background corrected using argon gas blanks, with the laser shutoff, which were measured 30 s before each ablation. These gas blankdata were also used for the calculation of the actual limits of detection(LODs) for the respective sample batch (n = 20). The following 40 swith laser ablation of the otolith target area were used for data acquisi-tion. To avoid overlapping signal intensities, another 30 s were used forthe monitoring of the decaying signal after the ablation was finished.The total acquisition time per analysis summed to 100 s time resolvedraw data. The software package LAMTrace (Simon Jackson, MacquarieUniversity, Sydney, Australia) was used for re-evaluation and quantifi-cation of the time-resolved raw data. To account for instrument drift,the analytical system was recalibrated every 16 ablations by analysinga NIST 610 (National Institute of Standards and Technology, Gaithers-burg, Maryland, USA) glass standard twice at the start and at the endof each sample batch (Longerich et al. 1996). In addition, the above-mentioned in-house Reference Otolith was analysed twice within everybatch of 14 unknown sample analyses. The position of the samplingcraters was viewed with a CCD camera and controlled by moving thesampling chamber on a three-directional stage.

The otolith samples from adult S. mentella were analysed in thenucleus region, along the juvenile zones (third-year growth) and alongthe marginal increments using the Agilent ICP-MS, while all other sam-ples were analysed in the nucleus region, within the marginal incrementsonly, using the Perkin-Elmer ICP-MS.A minimum of two replicate anal-yses were carried out in each otolith region. 42Ca was used as an internalstandard to correct for variations of the amount of ablated material,

45°

50°

55°

60°

65°

70°

75°

45°

50°

55°

60°

65°

70°

75°

�45° �15° 0° 15° 30°

S. marinusS. mentella

FC

WG

DS

FA

IR

EGIS

HeimlandRidge

BS

Greenland

ICESNAFO

�60° �30°

�45° �15° 0° 15° 30°�60° �30°

Fig. 1. Positions and areas of redfish otolith samples (see Table 1 forarea codes). Dashed lines represent boundaries between area definitions,the division between NAFO and ICES areas is indicated by a thickdashed line. Map source: GEBCO, Mercator projection.

assuming a constant concentration of calcium carbonate in the otolithmatrix. In this approach, the internal standard concentration was fixedat 40 wt% Ca (calculated as 56 wt% CaO). The NIST 610 was usedfor external calibration of all the other elements determined withinthe present study using 11.8 wt% CaO for internal standardisation. TheLODs were derived for each sample batch (20 analyses). With the Agi-lent ICP-MS, nine elements were quantified above the following averageLODs (µg g−1) throughout the analysis: Li: 0.06, Mg: 0.5, Mn: 0.08,Cu: 0.05, Zn: 0.1, Rb: 0.01, Sr: 0.5, Sn: 0.06 and Ba: 0.01. On thePerkin-Elmer ICP-MS, the average LODs (µg g−1) of the five exam-ined elements were: Li: 0.2, Mg: 1.0, Cu: 0.07, Sr: 0.01 and Ba: 0.01.The LODs for Mn (0.3 µg g−1) and Zn (0.5 µg g−1) were not suffi-cient for otolith analysis. Results for the Reference Otolith analysed


on both instruments showed good comparability within analytical error.Long-term reproducibility of LA-ICP-MS analyses, as estimated fromreplicates in the nucleus region of the Reference Otolith, were betterthan 15% for Rb, Sr, Ba and <40% for the remaining elements close todetection limits.

Statistical analysis

The average elemental concentrations were calculated (after normali-sation to Ca) from two or three sampling spots within the same region(nucleus, third-year annulus and edge) on each otolith. Because thesevalues were not normally distributed (Kolmogorov–Smirnov test fornormality, P < 0.05) for most elements, all concentrations were log10transformed before analysis. This transformation also accounted forthe largely different ranges of concentrations (e.g. Cu: 0.1–3.0 µg g−1,Sr: 1000–5000 µg g−1), decreasing the dominance of Sr in multivariateanalysis. Box-and-whisker plots of the elemental concentrations wereexamined to illustrate differences between sampling years and areas,as well as otolith regions. The significance of these comparisons wastested using multivariate analysis of variance (MANOVA) and separateANOVA for each element with year or area as categorical factors, fol-lowed by post hoc Tukey honest significant difference tests. Individualanalyses of covariance (ANCOVA) with fish total length as a covariatewere performed for each element to test for size effects on elementalconcentrations. To investigate possible relationships between elements,bivariate plots of the concentrations of all elemental pairs were examinedand tested for significance using a Pearson rank correlation.

The discrimination between geographic areas was tested for bothS. marinus and S. mentella using linear discriminant function analysis(LDFA). Scatterplots of the scores of the first two canonical variateswere drawn with 95% confidence ellipses to visualise area differences.The classification success between areas was determined by jackknifedcross-validation matrices (SYSTAT 9, SPSS Inc. 1999).

Results

Temporal variation

Before comparing otolith elemental signatures forS. marinus and S. mentella from various fishing areas inthe North Atlantic, the temporal stability of element con-centrations of juvenile redfish was investigated within alimited geographic area, namely the Heimland Ridge offEast Greenland. For both juvenile S. marinus (Fig. 2) andS. mentella (Fig. 3), year-to-year variation in the nucleusregion was relatively low, while larger differences betweenyears were observed in the marginal (edge) regions, espe-cially in case of Li and Ba. MANOVA of the elementalconcentrations in the nuclei with the sampling year as categor-ical factor indicated no significant differences between years(S. marinus: Pillai’s trace = 0.881, F15,36 = 0.89, P = 0.581;S. mentella: Pillai’s trace = 0.575, F10,42 = 1.69, P = 0.114),but significant year effects for the edge microchemistry (S.marinus: Pillai’s trace = 1.404, F15,30 = 1.76, P = 0.024; S.mentella: Pillai’s trace = 1.105, F10,42 = 5.18, P < 0.001).SeparateANOVA for each element showed no significant dif-ferences between years (P < 0.05) for the nucleus analyses,but significant year effects in the edge analyses of the S. mar-inus otoliths (Li: F3,13 = 10.05, P = 0.001; Cu: F3,12 = 4.01,P = 0.034) and the S. mentella otoliths (Li: F2,24 = 18.27,P < 0.001; Sr: F2,24 = 27.46, P < 0.001; Ba: F2,24 = 4.07,

P = 0.030). The post hoc Tukey tests indicated significantdifferences between the first sampling year (1998) and theconsecutive years for both species (Figs 2 and 3). Concentra-tions of Sr and Ba were consistently higher along the edgesthan in the nuclei, while Mg concentrations were always loweron the edge than in the nucleus region.

Geographic variation

The effect of fish size (total length), tested by individualANCOVA for each of the five elements consistently detectedin the otolith nuclei and edges of both species and in the juve-nile otolith zones of S. mentella (Li, Mg, Cu, Sr and Ba), wasnot significant (P > 0.05 in all cases).

Elemental concentrations in the otolith nucleus regionshowed highly significant differences between the geographicareas where the fish had been caught (MANOVA, P < 0.005,detailed results in Table 2) and geographic patterns for Liand Cu that repeated in both species (Figs 4 and 5): Li con-centrations were highest in the North-west Atlantic (FlemishCap) and North-east Atlantic (Barents Sea), whereas the cen-tral North Atlantic areas (Greenland, Iceland etc.) showedlower concentrations, with a consistent U-shaped trend fromWest to East. The opposite was found for Cu (and Mn inthe case of S. mentella), exhibiting highest concentrations inthe central areas. The observed geographic distribution pat-terns were also found in the edge regions of the analysedS. marinus otoliths, but were not present in the juvenile andedge regions of the S. mentella samples.

The Li concentrations measured in the Barents Sea sam-ples of adult S. marinus (Fig. 4) were significantly higher thanin all other areas (ANOVA; nuclei: F4,45 = 11.12, P < 0.001;edge: F4,43 = 31.19, P < 0.001; Tukey tests in both casesP < 0.01). Significant differences between areas were alsofound for Cu (ANOVA; nuclei: F4,39 = 4.12, P = 0.007;edge: F4,42 = 8.40, P < 0.001; Tukey tests in both casesP < 0.01) and for Ba in the nuclei (ANOVA; F4,45 = 2.76,P = 0.039; Tukey test n.s.). In all areas, Sr and Ba concentra-tions were consistently elevated in the edge region, comparedto the nucleus values, up to 2-fold in the case of the FlemishCap samples.

As in the S. marinus samples, the S. mentella otolithsfrom the Barents Sea showed higher Li values in the edgethan in the nucleus, while samples from all other areasexhibited decreased Li concentrations in the edge region(Fig. 5). Only the edge concentrations of Li, however, dif-fered significantly between areas (ANOVA; F6,45 = 9.87,P < 0.001). Cu, Rb and Sn showed significant geographic dif-ferences in the nuclei (ANOVA; Cu: F6,46 = 3.23, P = 0.010;Rb: F6,46 = 4.16, P = 0.002; Sn: F6,46 = 3.17, P = 0.011),whereas Zn, Sr and Sn in the edges were significantly dif-ferent between areas (ANOVA; Zn: F6,45 = 3.22, P = 0.010;Sr: F6,45 = 5.27, P < 0.001; Sn: F6,45 = 2.41, P = 0.042;Tukey tests only significant for Sr, P < 0.05). The post hocTukey tests indicated that these significances are mainly


Year

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

1.8A|A A|B A|A A|A

1998 1999 2000 2001 2002

1998 1999 2000 2001 2002

0

500

1000

1500

2000

2500

1998 1999 2000 2001 2002

1998 1999 2000 2001 2002

1998 1999 2000 2001 2002

0

2

4

6

8

10

12

14

16

18NucleusEdge

A|A A|B A|A A|AC

0

10

20

30

40

50

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

Mg

(�g/

g ot

olith

)Li

(�

g/g

otol

ith)

Ba

(�g/

g ot

olith

)S

r (�

g/g

otol

ith)

Cu

(�g/

g ot

olith

)

Year

Box plot legend

Non-outlier maximum

3rd quartile (75th percentile)

Median

1st quartile (25th percentile)

Non-outlier minimum

Fig. 2. Concentration (µg g−1 relative to Ca) of the five elements consistently measured above LOD in thenucleus and edge regions of juvenile S. marinus otoliths, collected on the Heimland Ridge/East Greenlandduring a period of five years (1998: n = 9; 2000: n = 3; 2001: n = 1; 2002: n = 5). The lower and upper marginof each box marks the first and third quartile respectively. The line between these represents the median, andthe lower and upper whiskers show the range to the minimum and maximum values respectively. Outliers aremarked by open circles, extreme values are indicated by stars. The results of the post hoc Tukey tests are givenabove the plots of those elements that showed significant differences in concentration between areas. Yearsthat do not share a common letter are significantly different.


1998 1999 2000 2001 20020

1

2

3

4

5

6

7A|A A|ABA|B

Year

1998 1999 2000 2001 20020

1

2

3

4

5

1998 1999 2000 2001 20020

4

8

12

16

20

24

NucleusEdge

A|A A|B A|BC

1998 1999 2000 2001 20020

10

20

30

40

50

60

70

Year

1998 1999 2000 2001 20020

500

1000

1500

2000

2500

3000A|A A|B A|C

Mg

(�g/

g ot

olith

)Li

(�

g/g

otol

ith)

Ba

(�g/

g ot

olith

)S

r (�

g/g

otol

ith)

Cu

(�g/

g ot

olith

)

Box plot legend

Non-outlier maximum


Median


Non-outlier minimum

Fig. 3. Concentration (µg g−1 relative to Ca) of the five elements consistently measured above the LOD inthe nucleus and edge regions of juvenile S. mentella otoliths, collected on the Heimland Ridge/East Greenlandduring a period of four years (n = 9 in 1998, 1999 and 2001). For box plot explanations, see Fig. 2.

driven by the differences between East Greenland, Ice-land and the Faroes. Further information on elementalconcentrations in different otolith regions was introduced bysampling the juvenile growth zones. The Mn concentrations

decreased from the nucleus over the third annulus to the edgein all areas, whereas Sr and Ba showed increasing trends fromnucleus to edge. Consistent otolith region trends within areaswere not observed for the other detected elements.


Table 2. MANOVA results for area effect in microchemistry of the analysed regions on the redfishotolith sections, based on log10-transformed element concentrations relative to Ca

Species Otolith region Pillai’s trace F Effect, error d.f. P

S. marinus Nucleus 1.034 2.65 20, 152 <0.0001S. marinus Edge 1.543 4.90 20, 156 <0.0005S. mentella Nucleus 1.845 2.07 54, 252 <0.0001S. mentella Juvenile 1.762 1.80 54, 234 <0.005S. mentella Edge 2.024 2.32 54, 246 <0.00001

Discriminant analysis of the log10-transformed concentra-tions of the elements that were consistently detectable in theotoliths (S. marinus: Li, Mg, Cu, Sr, Ba; S. mentella: Li, Mg,Mn, Cu, Zn, Rb, Sr, Sn, Ba), however, revealed only weakseparation between geographic areas. Discriminant functionscores for the S. marinus nucleus samples illustrate highsimilarity in elemental signatures between West and EastGreenland (Fig. 6a). For the edge signatures, the Greenlandareas and the Barents Sea separate from a group of samplesfrom Iceland and the Flemish Cap (Fig. 6b). The nucleussignatures of the S. mentella samples show considerablyless geographic discrimination, with some similarity betweenEast Greenland and the Irminger Sea, as well as betweenIceland and the Faroe Islands (Fig. 7a), whereas these asso-ciations were not apparent in the edge signatures (Fig. 7b).These patterns are confirmed by the cross-validation matri-ces (Tables 3–6), showing high misclassification rates(S. marinus: 11–44%, S. mentella: 10–30%) between thecentral North Atlantic areas. The overall classification suc-cess using nucleus signatures was poor for both species(S. marinus: 43%, Table 3; S. mentella: 46%, Table 4). Whencombining the central North Atlantic samples (East Green-land, Iceland, Irminger Sea and Faroe Islands) to one area, theoverall classification success increased to 64% for S. marinusand to 67% for S. mentella. The cross-validation based on theedge signatures of the S. marinus otoliths confirmed 100%separation of the Barents Sea samples (Table 5), whereas thehighest correct classification of the S. mentella samples wasonly 50% in the case of Iceland and the Barents Sea (Table 6).

Migration

Evidence was found during autumn 1998 and summer 1999,using tracking strong cohorts in the length–frequency distri-butions in these areas, for migration of juvenile S. mentellawith a size range of 25–30 cm total length from the EastGreenland shelf into the oceanic occurrences in the centralIrminger Sea (Stransky 2000). A subset of otoliths avail-able from this period and areas was analysed for traceelements in the nucleus and in the edge regions, reflect-ing the first-year and the recently inhabited environment.The nucleus microchemistry was very similar between EastGreenland and the Irminger Sea for all five detected elements

(Fig. 8). No significant differences in elemental signatureswere found between areas (MANOVA; Pillai’s trace = 0.578,F5,9 = 2.47, P = 0.113). The elemental composition in theedge regions, however, was significantly different betweenthe two geographic areas (MANOVA; Pillai’s trace = 0.616,F5,11 = 3.53, P = 0.038), indicated by lower concentrationsof Li, Mg and Ba in the Irminger Sea, compared to EastGreenland (Fig. 8). Separate ANOVA for each element, how-ever, did not reveal significant differences between areas(P > 0.05).

Discussion

The elemental concentrations found in redfish otoliths aresimilar to values reported for other North Atlantic fishspecies, such as cod (Campana et al. 2000) and bluefin tuna(Secor et al. 2002) in the North-west Atlantic and deep-waterblack scabbardfish in the North-east Atlantic (Swan et al.2003a), using solution-based bulk composition approaches.

When analysing geographic variation in otolith micro-chemistry, the estimation of the level of temporal variation ofthe elemental signatures within an area is crucial, becausea sample taken in one particular time period in a cer-tain area might not be representative for samples taken inthe same area during another period (Gillanders 2002). Inorder to test possible interannual effects, the otoliths chosenin the present study were collected within an importantredfish nursery area off East Greenland with consistentlyhigh abundance of juvenile S. marinus and S. mentella(Magnússon et al. 1988). The present investigation showsthat the analysed fish are very likely to have originated fromthe same spawning ground, because no significant differencesbetween sampling years were found for the elements detectedin the otolith nuclei. In contrast, the significant year-effectsin edge chemistry of these otoliths, mainly caused by largelyvarying Li, Sr and Ba concentrations, indicate either a dif-fering environment inhabited during the recent few monthsbefore capture or inter-annual variation in habitat chemistry.The low number of samples available for some years and pos-sible edge contamination effects, however, leave some doubton this interpretation of the edge results. Considering themarkedly higher level of variation in elemental concentra-tions of Li, Sr and Ba observed for adult S. marinus and


Area

0

1

2

3

4

5

6

7

AreaFC WG EG IS BS

FC WG EG IS BS

FC WG EG IS BS FC WG EG IS BS

FC WG EG IS BS

0.0

0.4

0.8

1.2

1.6

2.0

2.4AB|A AB|AB|AA|BAB|AB

0

10

20

30

40

50

60

NucleusEdge

A|A B|BA|AA|AA|A

0

10

20

30

40

50

60

70

80

0

1000

2000

3000

4000

5000

Mg

(�g/

g ot

olith

)Li

(�

g/g

otol

ith)

Ba

(�g/

g ot

olith

)S

r (�

g/g

otol

ith)

Cu

(�g/

g ot

olith

) Box plot legend

Non-outlier maximum


Median


Non-outlier minimum

Fig. 4. Concentration (µg g−1 relative to Ca) of the five elements consistently measured above the LOD inthe nucleus and edge regions of adult S. marinus otoliths, collected in five different areas in the North Atlantic(for area codes, see Table 1; for box plot explanations, see Fig. 2).

S. mentella from various geographic areas in the NorthAtlantic, especially in the otolith edge regions, the year-to-year signatures in the nucleus of juvenile redfish otoliths canbe regarded as relatively stable.

The most striking geographic patterns in elementalconcentrations were found for Li and Cu in otolith nucleiof adult fish of both species and for Mn in S. mentellaotoliths. Following a longitudinal trend, Li decreased in


Area

FC DS IR EG IS FA BS0

2

4

6

8

10

FC DS IR EG IS FA BS0.0

0.5

1.0

1.5

2.0

2.5A|A AB|AAB|AB|AAB|AAC|AAB|A

0

5

10

15

20

25Nucleus JuvenileEdge Outliers Extremes

A|A A|BA|AA|AA|AA|AA|A

5

15

25

35

45

FC DS IR EG IS FA BS


FC DS IR EG IS FA BS FC DS IR EG IS FA BS


0.0

0.5

1.0

1.5

2.0

2.5

3.0

0.0

0.1

0.2

0.3

AD|A AD|ACD|ABD|AA|AAD|AAD|A

FC DS IR EG IS FA BS0.4

0.6

0.8

1.0

1.2

1000

2000

3000

4000

5000

6000A|AB A|ACFA|ACFA|CEA|CEA|BDFA|AC

AreaFC DS IR EG IS FA BS

0

1

2

3

4

5

6

7

8

Mg

(�g/

g ot

olith

)Li

(�

g/g

otol

ith)

Ba

(�g/

g ot

olith

)S

r (�

g/g

otol

ith)

Mn

(�g/

g ot

olith

)

Box plot legend

Non-outlier maximum


Median


Non-outlier minimum

Cu

(�g/

g ot

olith

)Z

n (�

g/g

otol

ith)

Rb

(�g/

g ot

olith

)S

n (�

g/g

otol

ith)

Fig. 5. Concentration (µg g−1 relative to Ca) of the nine elements consistently measured above the LOD inthe nucleus, third annulus and edge regions of adult S. mentella otoliths, collected in seven different areas in theNorth Atlantic (for area codes, see Table 1; for box plot explanations, see Fig. 2).

the otoliths from the North-west Atlantic to the centralareas (Greenland, Irminger Sea, Iceland, Faroe Islands) andincreased from the central areas to the North-east Atlantic,while Mn and Cu showed opposite trends. Despite thesestrong signals and significant area effects for Li, Cu and Ba in

S. marinus otoliths and Cu, Rb and Sn in S. mentella otoliths,spatial separation derived from multivariate analysis wasrelatively poor. Attempts at using various sub-sets of ele-ments in the present study to improve discrimination betweenareas did not reveal higher classification success. A slight


BSISEGWGFC

Area

WG

EG

BS

IS

FC

�4 �3 �2 �1 0 1 2 3 4 5 6

CF1

�4

�3

�2

�1

0

1

2

3

4

5

6

CF

2

IS

FCBS

EG

WG

�3

�2

�1

0

1

2

3

�3 �2 �1 0 3

CF1

CF

2

BSISEGWGFC

Area

(a)

(b)

1 2

Fig. 6. Discriminant function scores and 95% confidence ellipses forS. marinus otoliths from five different areas in the North Atlantic (forarea codes, see Table 1), based on the log10-transformed concentrationsof five elements (Li, Mg, Cu, Sr, Ba) in the nucleus (a) and edge (b)region.

improvement of discriminatory power was accomplished bypooling the centralAtlantic areas to one group and comparingthis group with the North-west and North-east Atlantic areas.The S. marinus samples were better separated by the edgechemistry than by nucleus signatures, which indicates thatthis species experiences an increased degree of spatial seg-regation during the adult phase. Because the spatial dis-crimination of S. mentella samples by edge signatures wasconsiderably lower than derived from nucleus composition,this lack of separation by nucleus signatures could eitherbe due to limited variability of the environmental chemi-cal composition or due to a common origin. Low overallcross-validated classification rates of 45% were reportedfor other North Atlantic fish species dwelling in deeper

BSFAISEGIRDSFC

AreaFC

FA

IS

DS

BS

EGIR

�4

�3

�2

�1

0

1

2

3

4

5

6

7

8

BSFAISEGIRDSFC

Area

EG

BS

FCDS

IS

IRFA

�4

�3

�2

�1

0

1

2

3

4

5

6

CF

2C

F2

(a)

(b)

�4 �3 �2 �1 0 1 2 3 4 5 6

CF1

�4 �3 �2 �1 0 1 2 3 4 5 6 7 8

CF1

Fig. 7. Discriminant function scores and 95% confidence ellipses forS. mentella otoliths from seven different areas in the North Atlantic (forarea codes, see Table 1), based on the log10-transformed concentrationsof nine elements (Li, Mg, Mn, Cu, Zn, Rb, Sr, Sn, Ba) in the nucleus(a) and edge (b) region.

habitats (Swan et al. 2003a, 2003b). The present study,along with comprehensive studies on deep-sea fish(Gordon et al. 2001), represents one of the fewinvestigations carried out in oceanic and deep-sea areas. Inthese offshore areas, trace element composition only variesmarginally, whereas more distinct spatial patterns are to beexpected in estuarine and coastal areas where the majority ofinvestigations have been carried out so far (Campana 1999).

Swan et al. (2003b) observed elevated levels of Li in wholeotoliths of the deep-water macrourid Nezumia aequalis onthe Reykjanes Ridge (SW of Iceland), which they relatedto relatively high input of trace elements from hydrother-mal activity in this region. Several other studies found Li to


Table 3. Jackknifed classification matrix of the discriminant function analysis of theelemental signatures (log10-transformed concentrations of Li, Mg, Cu, Sr and Ba relativeto Ca) in the nucleus region of S. marinus otoliths from five different areas in the North

Atlantic (for area codes, see Table 1)The percentages (and the number of samples in parentheses) in rows represent theclassification into the areas given in columns. Total classification success is 43%

FC WG EG IS BS

FC 62.5 (5) 12.5 (1) 0.0 (0) 12.5 (1) 12.5 (1)WG 11.1 (1) 22.2 (2) 44.4 (4) 11.1 (1) 11.1 (1)EG 10.0 (1) 50.0 (5) 20.0 (2) 0.0 (0) 20.0 (2)IS 37.5 (3) 12.5 (1) 0.0 (0) 50.0 (4) 0.0 (0)BS 33.3 (3) 0.0 (0) 0.0 (0) 0.0 (0) 66.7 (6)

contribute to spatial separation, e.g. Milton and Chenery(2001), Gillanders (2002) and Rooker et al. (2003). In thepresent study, the highest Li concentrations were found inthe Barents Sea samples, being confirmed by multiple andindependent measurements. Because little is known aboutvariation in Li concentrations in the North Atlantic, no con-clusive explanation could be found for the relatively highlevels of Li in redfish otoliths from the Barents Sea, especiallyin the edge regions. Li, as well as Mg and Ba, are conservativeelements, directly proportional to salinity (Bruland 1983).However, a relationship between the concentrations of theseelements in redfish otoliths with salinity, extracted from theWorld Ocean Database (Conkright et al. 2002), could not beobserved.

The observed differences in the concentration of severalelements in nucleus, juvenile and edge zones of the anal-ysed redfish otoliths are most probably related to ontogeneticor growth rate changes, commonly found for Sr:Ca ratios(Campana 1999). Sr as well as Ba concentrations of theinvestigated redfish otoliths were consistently higher in theedge regions than in the nuclei, which has also been describedfor other marine fish species, such as English sole (Brown2003). The elevated levels of Sr and Ba determined insolution-based analysis of the three-year core of the redfishReference Otolith provided further indications of ontogeneticincrease of these elements in the otolith zones following thefirst year of growth. Contrarily, Li and Mn decreased from thenucleus to the marginal zones. Brown (2003) also observedhigher Li values in the cores than in the edges but could notfind a conclusive explanation for this phenomenon. Gordonet al. (2001) performed line scans across otoliths of deep-seafish and found a distinct increase in Mn concentration fromotolith edge to nucleus for roundnose grenadier in the RockallTrough. Elevated Mn levels in cores of clupeid otoliths werereported recently by Brophy et al. (2004) and were related toembryological development or spatial variation in the calciumcarbonate crystal structure.

Migration patterns based on otolith microchemistry haveprimarily been studied for fish moving between freshwater

and marine habitats, utilising the effect of changes in ambientsalinity on otolith Sr:Ca ratios (Campana 1999). Few studiesto date have investigated migration routes of marine fish inan open ocean environment using multi-element patterns.Thepresent study represents the first approach to track migrationof an important fish species in the North-east Atlantic. Largequantities of juvenile S. mentella of the size range 25–30 cmwere observed in groundfish surveys off East Greenland in1996–1998 but almost disappeared in 1999. A considerableproportion of this size group was observed in 1999 as recruitsin the Irminger Sea (Stransky 2000). The between-area sim-ilarity in elemental signatures in the otolith cores on the onehand would support a migration from the East Greenlandshelf into the pelagic habitat of the Irminger Sea, but couldalso be due to a lack of differences between the ambient waterchemistry of these two areas.

The highly migratory behaviour of redfish and unresolvedstock delineation, particularly of S. mentella, has repeatedlyled to problems in fisheries management (ICES 1998). Aspart of a multidisciplinary project, the present study investi-gated the use of otolith microchemistry for stock separationof S. marinus and S. mentella. The derived weak geographicseparation has also been observed in body morphometrics(Garabana Barro 2005) and otolith shape analysis (Stransky2005). Although these findings do not necessarily imply highlevels of gene flow, the results obtained for S. mentella arein accordance with the low heterogeneity of samples acrossthe North Atlantic reported in the genetic study of Roqueset al. (2002), based on microsatellite DNA. The moderateseparation of western, central and eastern areas, as derived inthese studies for both redfish species, generally supports thecurrent division of redfish management units (ICES 1998)on a large scale. In contrast, previous haemoglobin analy-ses indicated the existence of several stocks of S. mentellain the Irminger Sea and adjacent waters (Johansen et al.2000). The integration of the outcome of the present studywith concurrently undertaken genetic analyses and morpho-metric approaches will further elucidate the degree of stockstructure of North Atlantic redfish and will provide a more


Table 4. Jackknifed classification matrix of the discriminant function analysis of the elemental sig-natures (log10-transformed concentrations of Li, Mg, Mn, Cu, Zn, Rb, Sr, Sn and Ba relative to Ca)in the nucleus region of S. mentella otoliths from seven different areas in the North Atlantic (for area

codes, see Table 1)The percentages (and the number of samples in parentheses) in rows represent the classification into the

areas given in columns. Total classification success is 46%


FC 60.0 (3) 0.0 (0) 0.0 (0) 20.0 (1) 0.0 (0) 0.0 (0) 20.0 (1)DS 0.0 (0) 80.0 (4) 0.0 (0) 0.0 (0) 20.0 (1) 0.0 (0) 0.0 (0)IR 0.0 (0) 10.0 (1) 20.0 (2) 30.0 (3) 0.0 (0) 30.0 (3) 10.0 (1)EG 0.0 (0) 0.0 (0) 22.2 (2) 66.7 (6) 0.0 (0) 0.0 (0) 11.1 (1)IS 0.0 (0) 0.0 (0) 0.0 (0) 20.0 (2) 60.0 (6) 20.0 (2) 0.0 (0)FA 0.0 (0) 0.0 (0) 0.0 (0) 20.0 (2) 20.0 (2) 40.0 (4) 20.0 (2)BS 20.0 (1) 0.0 (0) 20.0 (1) 0.0 (0) 20.0 (1) 40.0 (2) 0.0 (0)

Table 5. Jackknifed classification matrix of the discriminant function analysis of the elemental sig-natures (log10-transformed concentrations of Li, Mg, Cu, Sr and Ba relative to Ca) in the edge region

of S. marinus otoliths from five different areas in the North Atlantic (for area codes, see Table 1)The percentages (and the number of samples in parentheses) in rows represent the classification into the


FC WG EG IS BS

FC 42.9 (3) 14.3 (1) 0.0 (0) 28.6 (2) 14.3 (1)WG 0.0 (0) 44.4 (4) 33.3 (3) 22.2 (2) 0.0 (0)EG 10.0 (1) 40.0 (4) 50.0 (5) 0.0 (0) 0.0 (0)IS 33.3 (3) 0.0 (0) 22.2 (2) 44.4 (4) 0.0 (0)BS 0.0 (0) 0.0 (0) 0.0 (0) 0.0 (0) 100.0 (10)

Table 6. Jackknifed classification matrix of the discriminant function analysis of the elemental sig-natures (log10-transformed concentrations of Li, Mg, Mn, Cu, Zn, Rb, Sr, Sn and Ba relative to Ca)in the edge region of S. mentella otoliths from seven different areas in the North Atlantic (for area

codes, see Table 1)The percentages (and the number of samples in parentheses) in rows represent the classification into the



FC 40.0 (2) 0.0 (0) 20.0 (1) 20.0 (1) 0.0 (0) 0.0 (0) 20.0 (1)DS 0.0 (0) 40.0 (2) 0.0 (0) 20.0 (1) 20.0 (1) 0.0 (0) 20.0 (1)IR 25.0 (2) 0.0 (0) 12.5 (1) 12.5 (1) 12.5 (1) 25.0 (2) 12.5 (1)EG 0.0 (0) 22.2 (2) 22.2 (2) 44.4 (4) 11.1 (1) 0.0 (0) 0.0 (0)IS 0.0 (0) 30.0 (3) 0.0 (0) 10.0 (1) 50.0 (5) 10.0 (1) 0.0 (0)FA 30.0 (3) 0.0 (0) 0.0 (0) 30.0 (3) 0.0 (0) 40.0 (4) 0.0 (0)BS 0.0 (0) 25.0 (1) 25.0 (1) 0.0 (0) 0.0 (0) 0.0 (0) 50.0 (2)

holistic basis for adaptive management of these valuableresources.

In conclusion, the interannual variation in elementalconcentrations in otolith nuclei of juvenile redfish in animportant nursery area on the East Greenland shelf wasnot significant, indicating temporal stability of elemental

signatures within a limited sampling area. Although con-sistent geographic patterns were observed for Li and Cuin otolith nuclei of both S. marinus and S. mentella(and additionally Mn in S. mentella otoliths) and sev-eral elements showed significant differences between sam-pling sites, only weak large-scale separation between


AreaEast Greenland

0

1

2

3

4

5

6

AreaEast Greenland Irminger Sea

0.0

0.4

0.8

1.2

1.6

2.0

2.4

East Greenland Irminger Sea0

5

10

15

20

25


10

20

30

40

50


500

1000

1500

2000

2500

3000

3500M

g (�

g/g

otol

ith)

Li (

�g/

g ot

olith

)

Ba

(�g/

g ot

olith

)S

r (�

g/g

otol

ith)

Cu

(�g/

g ot

olith

) Box plot legend

Non-outlier maximum


Median


Non-outlier minimum

Irminger Sea

NucleusEdge

Fig. 8. Concentration (µg g−1 relative to Ca) of the five elements consistently measured above LOD in thenucleus and edge regions of S. mentella otoliths from fish with total lengths of 25–30 cm, collected on the EastGreenland shelf in October 1998 and in the Irminger Sea in July 1999 during a large-scale migration observedfor these size groups, areas and period. For box plot explanations, see Fig. 2.

the western, central and eastern North Atlantic couldbe derived from multivariate analysis of the elementalsignatures. The elemental composition of the otolith edgeregions provided better separation of the S. marinusotoliths and decreased discrimination of the S. mentella

samples compared to the nucleus signatures. It is thereforeunclear if the lack of clear spatial separation from natalsignatures is due to limited variability of the chemical com-position of the ambient environment or due to a commonorigin.


Acknowledgements

We would like to express our gratitude toFran Saborido-Rey (Institute of Marine Research, Vigo,Spain), Margaret Treble (Fisheries and Oceans Canada, Win-nipeg, Manitoba), Thorsteinn Sigurdsson (Marine ResearchInstitute, Reykjavík, Iceland), Jakúp Reinert (Faroese Fish-eries Laboratory,Torshavn, Faroe Islands) and Kjell Nedreaas(Institute of Marine Research, Bergen, Norway) for provid-ing redfish otoliths.The Greenland and Irminger Sea sampleswere collected onboard the German FRV ‘Walther HerwigIII’ with the help of several staff members and volunteers.Manfred Stein (Federal Research Centre for Fisheries, Insti-tute for Sea Fisheries, Hamburg, Germany) introduced thefirst author to Ocean DataView, oceanographic databases andrecent knowledge on current patterns and water masses in theNorth Atlantic. Philip Yeats (Fisheries and Oceans Canada,Bedford Institute of Oceanography, Dartmouth, Canada)gave advice on trace element properties and distributionacross the North Atlantic, and Friedrich Nast (Bundesamt fürSeeschifffahrt und Hydrographie, Deutsches Ozeanographis-ches Datenzentrum, Hamburg, Germany) helped withcollating data on contaminant concentrations in the NorthAtlantic. Cornelius Hammer and Soenke Jansen gave helpfulcomments on the manuscript, further valuable suggestionswere made by Steve Campana and two anonymous referees.This work was partly funded by the European Commissionwithin the 5th Framework Programme, Specific Programme‘Quality of Life and Management of Living Resources’,Key Action 5: ‘Sustainable Agriculture, Fisheries andForestry’ (R&D project REDFISH, QLK5-CT1999-01222).Travel funds for the first author’s participation in the ‘ThirdInternational Symposium on Fish Otolith Research andApplication’, where this work was presented, were providedby Stiftung Seeklar and Hansische Universitätsstiftung.

References

Brophy, D., Jeffries, T. E., and Danilowicz, B. S. (2004). Elevatedmanganese concentrations at the cores of clupeid otoliths: possibleenvironmental, physiological, or structural origins. Marine Biology144, 779–786. doi:10.1007/S00227-003-1240-3

Brown, J. A. (2003). An evaluation of the nursery role of estuaries forflatfish populations in central California. Ph.D. Thesis, Universityof California Santa Cruz, CA.

Bruland, K. W. (1983). Trace elements in sea-water. In ‘ChemicalOceanography Vol. 8’. (Eds J. P. Riley and R. Chester.) pp. 157–220.(Academic Press: London.)

Campana, S. E. (1999). Chemistry and composition of fish otoliths:pathways, mechanisms and applications. Marine Ecology ProgressSeries 188, 263–297.

Campana, S. E., Thorrold, S. R., Jones, C. M., Günther, D., Tubrett, M.,et al. (1997). Comparison of accuracy, precision, and sensitivityin elemental assays of fish otoliths using the electron micro-probe, proton-induced X-ray emission, and laser ablation inductivelycoupled plasma mass spectrometry. Canadian Journal of Fish-eries and Aquatic Sciences 54, 2068–2079. doi:10.1139/CJFAS-54-9-2068

Campana, S. E., Chouinard, G. A., Hanson, J. M., Fréchet, A., andBrattey, J. (2000). Otolith elemental fingerprints as biological tracersof fish stocks. Fisheries Research 46, 343–357. doi:10.1016/S0165-7836(00)00158-2

Conkright, M. E., Antonov, J. I., Baranova, O., Boyer, T. P.,Garcia, H. E., et al. (2002). World Ocean Database 2001, Volume 1:Introduction. (Ed. S. Levitus.) NOAA Atlas NESDros. Inf. Serv. 42.US Government Printing Office, Washington, DC.

Elsdon, T. S., and Gillanders, B. M. (2003). Relationship betweenwater and otolith elemental concentrations in juvenile black breamAcanthopagrus butcheri. Marine Ecology Progress Series 260,263–272.

Garabana Barro, D. (2005). The genus Sebastes Cuvier, 1829 (Pisces,Scorpaenidae) in the north Atlantic: species and stock discrimina-tion using traditional and geometric morphometrics. Ph.D. Thesis,University of Vigo, Spain.

Gillanders, B. M. (2002). Temporal and spatial variability in elementalcomposition of otoliths: implications for determining stock identityand connectivity of populations. Canadian Journal of Fisheries andAquatic Sciences 59, 669–679. doi:10.1139/F02-040

Gordon, J. D. M., Swan, S. C., Geffen, A. J., and Morales-Nin, B.(2001). Otolith microchemistry as a means of identifying stocks ofdeep-water demersal fishes (OTOMIC). NAFO Scientific CouncilResearch Documents 01/100, 1–10.

Günther, D., and Heinrich, C. A. (1999). Enhanced sensitivityin LA-ICP-MS using helium-argon mixtures as aerosol car-rier. Journal of Analytical Atomic Spectrometry 14, 1363–1368.doi:10.1039/A901648A

ICES (1998). Report of the Study Group on Redfish Stocks. ICESC.M. 1998/G:3. International Council for the Exploration of the Sea,Copenhagen.

Johansen, T., Daníelsdóttir, A. K., Meland, K., and Nævdal, G. (2000).Studies of the genetic relationship between deep-sea and oceanicSebastes mentella in the Irminger Sea. Fisheries Research 49,179–192. doi:10.1016/S0165-7836(00)00193-4

Longerich, H. P., Jackson, S. E., and Günther, D. (1996). Laser ablationinductively coupled plasma mass spectrometric transient signal dataacquisition and analyte concentration calculation. Journal of Analyt-ical Atomic Spectrometry 11, 899–904. doi:10.1039/JA9961100899

Magnússon, J., and Magnússon, J. V. (1995). Oceanic redfish (Sebastesmentella) in the Irminger Sea and adjacent waters. Scientia Marina59, 241–254.

Magnússon, J., Kosswig, K., and Magnússon, J. V. (1988).Young redfishon the nursery grounds in the East Greenland shelf area. ICES C.M.1988/G:38. International Council for the Exploration of the Sea,Copenhagen.

Marcogliese, D. J.,Albert, E., Gagnon, P., and Sévigny, J.-M. (2003). Useof parasites in stock identification of the deepwater redfish (Sebastesmentella) in the Northwest Atlantic. Fishery Bulletin 101, 183–188.

Milton, D. A., and Chenery, S. R. (2001). Can otolith chemistry detectthe population structure of the shad hilsa Tenualosa ilisha? Compar-ison with the results of genetic and morphological studies. MarineEcology Progress Series 222, 239–251.

Nedreaas, K., Johansen, T., and Nævdal, G. (1994). Geneticstudies of redfish (Sebastes spp.) from Icelandic and Green-land waters. ICES Journal of Marine Science 51, 461–467.doi:10.1006/JMSC.1994.1047

Roques, S., Sévigny, J.-M., and Bernatchez, L. (2002). Geneticstructure of deep-water redfish, Sebastes mentella, popula-tions across the North Atlantic. Marine Biology 140, 297–307.doi:10.1007/S002270100705

Rooker, J. R., Secor, D. H., Zdanowicz, V. S., de Metrio, G.,and Relini, L. O. (2003). Identification of Atlantic bluefin tuna(Thunnus thynnus) stocks from putative nurseries using otolith


chemistry. Fisheries Oceanography 12, 75–84. doi:10.1046/J.1365-2419.2003.00223.X

Saborido-Rey, F. (1994). The genus Sebastes Cuvier, 1829 (Pisces,Scorpaenidae) in the NorthAtlantic: species and population identifi-cation using morphometric techniques; growth and reproduction ofthe Flemish Cap populations. Ph.D. Thesis, Universidad AutónomaMadrid.

Saborido-Rey, F., and Nedreaas, K. H. (2000). Geographic variationof Sebastes mentella in the Northeast Arctic derived from a mor-phometric approach. ICES Journal of Marine Science 57, 965–975.doi:10.1006/JMSC.2000.0703

Secor, D. H., Campana, S. E., Zdanowicz, V. S., Lam, J. W. H., Yang, L.,and Rooker, J. R. (2002). Inter-laboratory comparison ofAtlantic andMediterranean bluefin tuna otolith microconstituents. ICES Journalof Marine Science 59, 1294–1304. doi:10.1006/JMSC.2002.1311

SPSS Inc. (1999). ‘SYSTAT® 9 Statistics I.’ (SPSS Inc.: Chicago, IL.)Stransky, C. (2000). Migration of juvenile deep-sea redfish (Sebastes

mentella Travin) from the East Greenland shelf into the centralIrminger Sea. ICES C.M. 2000/N:28. International Council for theExploration of the Sea, Copenhagen.

Stransky, C. (2005). Geographic variation of golden redfish (Sebastesmarinus) and deep-sea redfish (S. mentella) in the North Atlanticbased on otolith shape analysis. ICES Journal of Marine Science 62,in press.

http://www.publish.csiro.au/journals/mfr

Swan, S. C., Gordon, J. D. M., and Shimmield, T. (2003a). Prelimi-nary investigations on the uses of otolith microchemistry for stockdiscrimination of the deep-water black scabbardfish (Aphanopuscarbo) in the North East Atlantic. Journal of Northwest AtlanticFisheries Science 31, 221–231.

Swan, S. C., Gordon, J. D. M., Morales-Nin, B., Shimmield, T.,Sawyer, T., and Geffen, A. J. (2003b). Otolith microchemistryof Nezumia aequalis (Pisces: Macrouridae) from widely dif-ferent habitats in the Atlantic and Mediterranean. Journal ofthe Marine Biological Association of the UK 83, 883–886.doi:10.1017/S0025315403007987H

Thresher, R. E. (1999). Elemental composition of otoliths as astock delineator in fishes. Fisheries Research 43, 165–204.doi:10.1016/S0165-7836(99)00072-7

Whitehead, P. J. P., Bauchot, M.-L., Hureau, J.-C., Nielsen, J., andTortonese, E. (Eds) (1986). ‘Fishes of the north-eastern Atlantic andthe Mediterranean (Poissons de l’Atlantique du nord-est et de laMediterranee).’ Volume III. (UNESCO: Paris.)

Yoshinaga, J., Nakama, A., Morita, M., and Edmonds, J. S. (2000). Fishotolith reference material for quality assurance of chemical analyses.Marine Chemistry 69, 91–97. doi:10.1016/S0304-4203(99)00098-5

Manuscript received 13 July 2004; revised 4 April 2005; and accepted12 April 2005.