S

Ia

JTa

b

a

ARRA

KMMOOT

1

mceeatB22a

eT

0h

Fisheries Research 148 (2013) 56– 63

Contents lists available at ScienceDirect

Fisheries Research

j our nal ho mep ag e: www.elsev ier .com/ locate / f i shres

hort communication

nsights into mixing and movement of South Pacific albacore Thunnuslalunga derived from trace elements in otoliths

ed I. Macdonalda,∗, Jessica H. Farleyb, Naomi P. Clearb, Ashley J. Williamsa,hor I. Carterb, Campbell R. Daviesb, Simon J. Nicola

Oceanic Fisheries Programme, Secretariat of the Pacific Community, BPD5-98848 Nouméa Cedex, New CaledoniaWealth from Oceans Flagship, CSIRO Marine and Atmospheric Research, GPO Box 1538, Hobart, Tasmania 7001, Australia

r t i c l e i n f o

rticle history:eceived 22 February 2013eceived in revised form 1 August 2013ccepted 8 August 2013

eywords:obilityigrationcean frontstolith chemistryuna

a b s t r a c t

Information on the movement and stock structure of commercially important tunas underpins the effec-tive management of exploited populations. In the case of the South Pacific albacore (Thunnus alalunga)stock, longstanding questions remain regarding the degree of connectivity among larval pools, the migra-tion routes of juveniles and adults and the biophysical factors influencing these processes. We measuredtrace elements (Li, Mg, Mn, Cu, Sr, Ba, Pb, Ca) in albacore otoliths collected across a broad geographicalrange in the South Pacific Ocean to address these knowledge gaps. Capture locations in French Polyne-sia, New Caledonia and New Zealand were discriminated with high accuracy (overall 85% of individualscorrectly classified) based on analyses at the otolith edge (reflecting the final <1 month of life) usingLA-ICPMS. Spatial comparisons of otolith core chemistry (reflecting the first ∼2 weeks of life post-hatch)from the 2005/06 cohort suggest some mixing of larval pools for fish sampled from New Caledonia andNew Zealand, whereas French Polynesian fish may have originated from a chemically and/or geograph-ically distinct larval source. Annual and/or sub-annual cycles in Sr:Ca and Ba:Ca were evident alongablation transects encompassing the full life history of individuals. These patterns may reflect seasonal

north-south movements across ocean fronts; however, the vertical behaviours of albacore and the lackof opportunities for controlled experiments on temperature effects and time-lags in elemental incor-poration complicates environmental reconstructions based on trace element data alone. Expanding thepresent analysis across multiple years and regions, and integrating data from several sources (e.g. com-mercial catch data, tag returns, otolith �13C and �18O, ocean circulation models) could help clarify thelinkages between environmental factors and mixing and movement patterns in albacore.

. Introduction

Tunas are highly mobile fishes that often undertake long-rangeovements to track prey resources, both horizontally and verti-

ally, and to reproduce at distant spawning grounds (Pattersont al., 2008; Rooker et al., 2008; Block et al., 2011). Mark-recapturexperiments using conventional tags, and data from satellite,coustic and archival tagging programmes have provided impor-ant insights into the movements of tunas (Dagorn et al., 2000;lock et al., 2001; Sibert and Hampton, 2003; Patterson et al.,

008), including albacore Thunnus alalunga (Ichinokawa et al.,008; Childers et al., 2011). However, unlike the North Pacificnd North Atlantic Oceans and the Mediterranean Sea where such

∗ Corresponding author. Present address: Faculty of Life and Environmental Sci-nces, University of Iceland, Sturlugata 7, 101 Reykjavík, Iceland.el.: +354 770 0662.

E-mail address: jedimacdonald@gmail.com (J.I. Macdonald).

165-7836/$ – see front matter © 2013 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.fishres.2013.08.004

© 2013 Elsevier B.V. All rights reserved.

tagging methods have been successfully applied to albacore (e.g.Otsu and Uchida, 1963; Bertignac et al., 1999; Arrizabalaga et al.,2002; Ichinokawa et al., 2008; Childers et al., 2011), similar resultshave not been observed in the South Pacific, with only smallnumbers of tags recovered (Labelle, 1993; Bertignac et al., 1996;Williams et al., 2010). This in part has been due to the difficulties intagging sufficient numbers of juveniles in the surface fisheries and,until quite recently, low exploitation rates in the longline fisheries(Williams et al., 2010). Moreover, the poor condition of many adultfish captured via longline methods has often prevented tagging, orresulted in high tagging-induced mortality (Williams et al., 2010).

Associating long-term catch per unit effort (CPUE) datasetswith trends in oceanographic parameters has proved useful inshowing how climatic variability (Kimura et al., 1997; Lu et al.,1998; Briand et al., 2011), localised environmental forces (Laurs

et al., 1977; Zainuddin et al., 2008; Domokos, 2009; Lan et al.,2012), prey distribution (Domokos et al., 2007) or combinationsof these factors (Briand et al., 2011) can influence the horizontaland vertical distribution, movement and recruitment dynamics of

ries Research 148 (2013) 56– 63 57

anbfts

atgalNspda2nise2

aiaactfieawi

2

2

l(((2tcu2aFsw

(ufivd

e(

Table 1Summary of capture locations, fish numbers (n), mean fork length (FL) and ranges(in parentheses), estimated ages from annual increment counts (from Farley et al.,2013a) and season spawned for albacore used in this study.

Capture location n FL (mm) Age class Seasonspawned

New Caledonia 1 84.0 2+ 2007/086 82.8 (81–85) 3+ 2006/075 83.4 (81–87) 3+ 2005/06

New Zealand 6 81.9 (81–83) 3+ 2006/072 82.5 (80–85) 4+ 2005/06

French Polynesia 4 91.3 (88–93) 3+ 2005/063 89.3 (86–91) 4+ 2004/05

J.I. Macdonald et al. / Fishe

lbacore. However, fundamental questions remain regarding theumber and/or location of spawning areas, connectivity and mixingetween larval sources and adult populations and the biophysicalorces that structure these connections. This situation has limitedhe spatial resolution of the population model used to assess thetatus of the stock (Kolody et al., 2010; Hoyle, 2011).

Analysis of the chemical constituents of otoliths may providen alternative method for elucidating mixing and movement pat-erns in albacore. Otoliths can be considered as natural tags thatrow continuously throughout the fish’s life. They form daily andnnual growth bands, and are composed of a calcium carbonateattice that is not re-metabolised once deposited (Campana andielson, 1985; Campana, 1999). These properties, combined with

trong evidence that certain trace elements and isotopes are incor-orated into otoliths in response to changes in environmental,ietary, physiological and/or ontogenetic parameters (e.g. Elsdonnd Gillanders, 2003; Macdonald and Crook, 2010; Walther et al.,010), have allowed scientists to trace movements and explore con-ections between larval and adult populations of marine fishes,

ncluding several tuna species, at a range of spatial and temporalcales (e.g. Swearer et al., 1999; Rooker et al., 2001, 2008; Wangt al., 2009; Schloesser et al., 2010; Shiao et al., 2010; Wells et al.,012).

Here we examine variation in trace elements in albacore otolithscross a large spatial scale spanning ∼19◦ latitude and 49◦ longituden the South Pacific Ocean to test the utility of otolith chemistrynalysis as a tool for defining movements and stock structure inlbacore. Specifically, we address the following questions: (1) Canapture locations be delineated based on the trace element concen-rations at the otolith edge? (2) Is there evidence for larval mixingollowing spawning, or for the existence of non-mixing populationsn the South Pacific? Additionally, we present examples of tracelement transects measured across the entire lifetimes of eightlbacore, and discuss the potential of combining such informationith tagging efforts and other isotopic, genetic and catch data to

ncrease inference in tracing movement patterns for the species.

. Materials and methods

.1. Otolith selection and preparation

Sagittal otoliths were sourced from albacore captured byongline vessels operating in waters off French Polynesia20◦13′ S–20◦38′ S, 146◦40′ W–146◦52′ W), New Caledonia21◦43′ S–23◦16′ S, 164◦10′E–165◦19′ E) and New Zealand39◦25′ S–39◦44′ S, 178◦24′ E–178◦27′ E) between September009 and May 2010. Fork length (FL) for each fish was measured tohe nearest cm and otoliths were removed immediately followingapture. Otoliths were cleaned of adhering tissue, washed in anltrasonic bath with deionised water then dried at 30 ◦C for at least4 h before being archived in polyethylene capsules. Otoliths from

subsample of 29 fish (New Caledonia, n = 12; New Zealand, n = 8;rench Polynesia, n = 9) with similar FL and capture date wereelected for analysis to maximise the likelihood that individualsould belong to the same cohort (Table 1).

One otolith from each fish was embedded in epoxy resinEpofix®, Struers) and sectioned transversely through the coresing a modified high-speed diamond cutting saw (GemastaTM)

tted with a 100-�m wide diamond impregnated blade. Trans-erse sections (∼1-mm thick) were mounted on a circular glassisc using araldite and polished to expose (assumed1) daily growth

1 Daily increment deposition has not been empirically validated for otoliths ofarly juvenile South Pacific albacore (Farley et al., 2013a). However, Laurs et al.1985) did demonstrate daily increment formation in North Pacific albacore sagittae

1 90.0 5+ 2003/041 95.0 7+ 2001/02

increments near the primordium using 1000× wet and dry sandpa-per and 3-�m lapping film. Sections were then turned over and theprocedure repeated on the opposite side. Prepared sections weretriple-rinsed in deionised water, air dried overnight in a class-100laminar flow cabinet, and arranged in rows on microscope slidesusing double-sided tape.

2.2. Trace element analysis

Otoliths were analysed using laser ablation-inductively cou-pled plasma mass spectrometry (LA-ICPMS). We used an Agilent7700 ICPMS (Agilent Technologies), coupled to a HelEx laser system(Laurin Technic, Canberra, and the Australian National Univer-sity) located at the School of Earth Sciences, The University ofMelbourne, Australia. The HelEx system is constructed around aCompex 110 (Lambda Physik, Gottingen, Germany) ArF excimerlaser (Eggins et al., 1998). Otolith mounts were placed in the samplecell and the primordium of each otolith was located visually with a400× objective and a video imaging system. The intended ablationpath on each sample was then digitally plotted using GeoStar v6.14software (Resonetics, USA).

Each otolith was ablated along a transect running from the pri-mordium along the antisulcul margin to the first inflection point(IP), then along the ventral arm, distal to the ventral groove, tothe terminal edge. A pre-ablation step was implemented to min-imise potential surface contamination (Proctor and Thresher, 1998;Davies et al., 2011), with the laser scanned at 12 �m s−1 across thesample at low energy using a 71 �m diameter spot. The resultingtransect was 6-�m deep based on calculations incorporating thelaser drilling rate, scan speed and power density on the sample forthe HelEx system in addition to microscopic examination of abla-tion site geometry (Eggins et al., 1998;Woodhead et al., 2005). Datawere then acquired from a second analysis along the same transectusing a 32 �m spot, with the laser pulsed at 10 Hz and scannedat 6 �m s−1 with fluence of ∼5 J cm−2. This produced a 32-�mwide × 6-�m deep transect, which, when daily increment pos-itions are considered for our samples, negates concerns regardingvertical versus horizontal growth integration errors (Hoover andJones, 2013). Ablation occurred inside a sealed chamber in an atmo-sphere of pure He (flow rate, ∼0.3 L min−1) with the vaporisedmaterial transported to the ICP-MS in the Ar carrier gas (flow rate,∼1.23 L min−1) via a signal smoothing manifold. Otoliths were ana-

7 25 43 55 63

lysed for a suite of elements including Li, Mg, Ca, Mn, Cu,88Sr and 138Ba, 207Pb and Ca was used as an internal standard tocorrect for variation in ablation yield among samples.

for fish between 50 and 100 cm FL, and suggested that increments are accreted dailyfrom the time of yolk-sac absorption in the species.

5 ries Research 148 (2013) 56– 63

swaTsritd3rM9S

2

egr2w2tatmgcptawop

dopf(rbpttt

2

ovrDtTsmtbecw

Table 2Univariate differences in each trace element concentration among capture locationsbased on analyses at the otolith edge (ANOVAs) and the otolith core (Kruskal–Wallistests). Analyses were performed on natural log transformed data. Significant differ-ences among locations ( = 0.05) are shown in bold.

Otolith edge Otolith core

d.f. MS F p H p

Li:CaLocation 2 0.183 1.254 0.304 0.273 >0.05Error 23 0.146

Mg:CaLocation 2 2.576 7.166 0.004 2.973 >0.05Error 23 0.359

Mn:CaLocation 2 0.378 2.643 0.093 0.914 >0.05Error 23 0.143

Cu:CaLocation 2 1.000 2.529 0.102 0.041 >0.05Error 23 0.395

Sr:CaLocation 2 0.012 1.656 0.213 1.814 >0.05Error 23 0.007

Ba:CaLocation 2 0.464 2.368 0.116 5.300 <0.048Error 23 0.196

Pb:Ca

8 J.I. Macdonald et al. / Fishe

Data processing was done offline using the Iolite version 2.13oftware (Paton et al., 2011). Subtraction of background ion countsas followed by the normalisation of each element to Ca using

n external calibration standard (National Institute of Standardsechnology, NIST612) which was analysed after every 10 otolithamples. Trace element concentrations were expressed as molaratios to Ca (i.e. Mg:Ca). We exported data at 1-s intervals, resultingn a spatial resolution along the time series of 6-�m horizontal dis-ance per integration. Measurement precision (% relative standardeviation – %RSD) was determined based on analyses of MACS-

(United States Geological Survey) (n = 8) and NIST610 (n = 8)eference standards run concurrently with the otolith samples.ean %RSD for the MACS-3 and NIST610 respectively was Li:Ca:

.42/7.26, Mg:Ca: 4.17/2.92, Mn:Ca: 3.80/3.66, Cu:Ca: 8.16/7.54,r:Ca: 1.93/0.64, Ba:Ca: 2.26/0.76 and Pb:Ca: 8.09/6.63.

.3. Age-related information and data selection

We estimated annual age and back-calculated birth year forach fish using validated methods based on counts of annualrowth zones in the sectioned otoliths (Farley et al., 2013a). Fishanged in age from 2.5 to 7.9 years and were spawned in the001/02–2006/07 summer spawning seasons. The majority of fishere in the 3+ or 4+ age classes and were spawned in either the

005/06 or 2006/07 austral summers (Table 1). The distance fromhe primordium to the otolith edge was measured along the laserblation transect for each otolith. In addition, distances betweenhe outer edges of the two most recently completed opaque (sum-

er/autumn) growth zones were used as an estimate of otolithrowth for the 12 months preceding capture. For eight otoliths withlear annual increments, measurements were also taken from therimordium to the outer margin of each opaque growth zone alonghe ablation transect. Measurements were made using an image-nalysis system (AnalySIS 3.2 Soft Imaging System) and imagesere acquired with an Olympus F-view II digital camera mounted

n a Leica Wild stereo microscope that was connected to a com-uter.

From these measurements we determined that the last 10 �m ofata in the horizontal plane at the terminal otolith edge representedtolith material accreted during the final two to four weeks of liferior to capture (otolith edge). Based on otolith size at hatch dataor albacore in the Mediterranean (García et al., 2006) and counts ofassumed) daily increments in South Pacific albacore otoliths (Sec-etariat of the Pacific Community, unpublished data), a data sliceetween 10 and 200 �m from the primordium in the horizontallane was selected to represent otolith material accreted duringhe first ∼2 weeks of life post-hatch (otolith core). For each fish,race element concentrations were averaged across each data sliceo produce an otolith edge and an otolith core value.

.4. Statistical analysis

We explored variation in trace element concentrations at thetolith edge among the three capture locations using one-way uni-ariate and multivariate analysis of variance (ANOVA and MANOVAespectively), and quadratic discriminant function analysis (QDFA).ata for all elements were natural log transformed to meet assump-

ions of normality and equality of variance-covariance matrices.ukey’s HSD pairwise comparisons were performed following aignificant ANOVA result and Pillai’s trace statistic was used forultivariate tests. We determined which multivariate combina-

ion of elements optimised discrimination success for the QDFA

ased on an iterative procedure similar to that described by Merciert al. (2011). Classification success was calculated by jackknifeross-validation matrices and standardised coefficients for the DFsere used to measure which elements were most important in

Location 2 1.109 0.970 0.394 2.291 >0.05Error 23 1.143

discriminating capture locations. Randomisation tests were usedto determine if the jackknife classification estimates were signifi-cantly different from random (White and Ruttenberg, 2007).

We chose to include only age 3+ and age 4+ individuals for theotolith edge analyses (see Table 1) to minimise potential for onto-genetic effects on elemental incorporation to bias our estimates ofvariation among capture locations (de Pontual et al., 2003;Waltheret al., 2010). No measurable difference in otolith edge chemistrywas detected between the two age-classes within each capturelocation (paired t-tests, all p > 0.1), and hence, we deemed it appro-priate to pool these data for analyses. As spawning and larval sourcelocations were unknown, we restricted our analyses at the otolithcore to fish spawned during the 2005/06 summer, and used uni-variate Kruskal–Wallis tests to assess spatial variability in otolithcore chemistry within this cohort. Sample sizes were small (i.e.French Polynesia, n = 4; New Caledonia, n = 5; New Zealand, n = 2),so we compared the Kruskal–Wallis statistic (H) to tabled criticalvalues to determine the significance level of each test, rather thanusing a Chi-square approximation. A Mann–Whitney U-test wasused to examine pairwise differences if H was significant. All statis-tical analyses were run in R 2.12.2 (The R Foundation for StatisticalComputing).

3. Results

3.1. Discrimination of capture locations

Variation among capture locations was observed in some, butnot all trace elements measured at the otolith edge (Fig. 1, Table 2).Significant spatial differences were present only for Mg:Ca, withsamples from French Polynesia displaying significantly highermean Mg:Ca (Tukey’s HSD, both p < 0.05) and enriched Ba:Ca

compared with those from New Caledonia and New Zealand. Con-centrations of Mg:Ca, Mn:Ca and Cu:Ca were lowest in the NewZealand samples, but differences among locations were not signif-icant. Low variability in Sr:Ca was seen across all locations (Fig. 1).

J.I. Macdonald et al. / Fisheries Research 148 (2013) 56– 63 59

Mn:C

a(µ

mol m

ol-1

)

0

1

2

3

4

0

1

2

3

4

5

0.0

0.5

1.0

1.5

2.0

2.5

0.0

0.5

1.0

1.5

Mg:C

a(m

molm

ol-1

)

0

1

2

3

4

5

Edge

0123456789

LI:C

a(µ

mo

l m

ol-1

) Core

0

2

4

6

8

Cu:C

a(µ

mol m

ol-1

)S

r:C

a(m

molm

ol-1

)

Ba:C

a( µ

mol m

ol-1

)P

b:C

a(µ

mol m

ol-1

)

Discriminant function 1

Dis

crim

inant

function 2

-4 -2 0 2 4

-3

-2

-1

0

1

2

3

NZ

NC

FP

Edge Core

FP NC NZ

FP NC NZFP NC NZ

FP NC NZ

Fig. 1. Mean (±SE) trace element concentrations measured at the otolith edge (open bars) and otolith core (shaded bars) for albacore captured from French Polynesia (FP),New Caledonia (NP) and New Zealand (NZ). The discriminant function plot shows spatial variation in multi-element signatures (Li:Ca, Mg:Ca, Mn:Ca, Cu:Ca, Sr:Ca, Ba:Ca)a lbacora

oMFcCfit7avfr

mong the three capture locations based on otolith edge data from age 3+ and 4+ are individual fish. Analyses are based on natural log transformed data.

Capture locations could be clearly discriminated based on antolith edge signature comprising six trace element ratios (Li:Ca,g:Ca, Mn:Ca, Cu:Ca, Sr:Ca, Ba:Ca–MANOVA, Pillai’s trace = 0.869,

= 2.435, d.f. = 12, 38, p = 0.018; Fig. 1). Overall, 85% of fish wereorrectly classified to their capture location. Fish captured in Newaledonia were classified with 100% accuracy. Small numbers ofsh caught in French Polynesia and New Zealand were misclassifiedo New Caledonia, reducing accuracy for those locations (71% and5% respectively). Discrimination was driven primarily by Mg:Ca

long the first DF, which explained 62.8% of the among-locationariance, and Ba:Ca and Mn:Ca on the second DF which accountedor the remaining dispersion in the data. The classification successate was significantly different from random (p < 0.01).

e. Ellipses represent 95% CIs around the centroid for each location, and data points

3.2. Otolith core chemistry

Values for most trace elements measured in otolith cores wererelatively invariant across the three capture locations (Fig. 1,Table 2). Samples from New Zealand and New Caledonia werenot significantly different in any of the comparisons. New Zealandsamples exhibited significantly higher mean Ba:Ca concentrationsthan those from French Polynesia (Mann–Whitney U test, p < 0.05),whilst also having the highest Li:Ca, Mn:Ca, Sr:Ca values among

the three locations (Fig. 1). French Polynesian samples were high-est in Mg:Ca, yet there was substantial variability in this markerwithin this location. Although direct comparisons between otolithcore and edge chemistry within each location were not the focus of

60 J.I. Macdonald et al. / Fisheries Research 148 (2013) 56– 63

0

1

2

3

4

0 50 0 100 0 150 0 2000 250 0 300

0

1

2

3

4

0 50 0 100 0 150 0 2000 250 0 300

0

1

2

3

4

0 50 0 100 0 150 0 2000 250 0 300

0

1

2

3

4

5

0

2

4

6

8

10

0 500 10 00 150 0 200 0 25 00 3000

0

2

4

6

8

10

12

14

16

18

0 500 10 00 150 0 200 0 25 00 3000

0

2

4

6

8

10

0 500 10 00 150 0 200 0 25 00 3000

0

2

4

6

8

10

1

2

3

4

1

2

3

4

1

3

2

1

23 4

5

1

2 34

1

2

3

4

5 6

78

1

2

3

4

12

3

4

ba

dc

fe

hg

10

8

6

4

2

0

10

8

6

4

2

0

10

8

6

4

2

0

10

8

6

4

2

0

10

8

6

4

2

0

10

8

6

4

2

0

18

16

14

12

10

8

6

4

2

0

18

16

14

12

10

8

6

4

2

0

4

3

2

1

0

4

3

2

1

0

4

3

2

1

0

4

3

2

1

0

4

3

2

1

0

4

3

2

1

0

5

4

3

1

2

0

5

4

3

1

2

0

0 50 0 1000 1500 2000 30002500 0 50 0 100 0 150 0 200 0 30002500

Distance from core (µm )

Ele

ment:C

a(m

mo

lm

ol-1

) Ele

ment:C

a(µ

mo

l mo

l -1)

Fig. 2. Trace element concentrations measured along transects ablated from the otolith core to the terminal edge along the ventral arm for eight representative samplescaptured from New Caledonia (a, b, c); French Polynesia (d, e, f); and New Zealand (g, h). The position of the first inflection point for each sample is overlaid (dashed verticall owth

( -handl o the r

ocS

3

z(Icm

ines) with annual growth increment locations (i.e. outer margin of each opaque grblack line) and Mg:Ca (blue line) are presented in mmol mol−1 and relate to the leftine), Ba:Ca (red line) and Pb:Ca (yellow line) are shown in �mol mol−1 and relate t

ur analysis, mean concentrations of Li:Ca, Mg:Ca and Mn:Ca wereonsistently higher at the otolith core versus the otolith edge, whiler:Ca and Ba:Ca values were always lower (Figs. 1 and 2).

.3. Patterns along trace element transects

Marked cycling of Sr:Ca was observed in the outer growthones of all otoliths analysed (Fig. 2). Minimum Sr:Ca values

∼1.4–1.6 mmol mol−1) were observed at, or close to, the firstP in all samples, and apart from some enrichment near theore, otolith Sr:Ca was relatively low and stable in the first fewonths of life (Fig. 2). Maximum values of the Sr:Ca ‘peaks’ ranged

zone along the ablation transect) denoted by grey arrows. Concentrations of Sr:Ca vertical axis. Concentrations of Li:Ca (green line), Mn:Ca (purple line), Cu:Ca (greyight-hand vertical axis.

between ∼2.5 and 4.3 mmol mol−1, with the concentration changebetween maxima and minima in adjacent ‘troughs’ in the range of1–2 mmol mol−1. Annual periodicity in Sr:Ca cycling was evidentin some cases (e.g. Fig. 2f); however, multiple cycles sometimesoccurred within a year (Fig. 2a and c). Sr:Ca peaks were often coin-cident with the position of the of opaque growth zones (e.g. Fig. 2b,c, f and g) but this was not always the case. Annuli in some sampleswere associated exclusively with troughs in Sr:Ca (e.g. Fig. 2e and

h) or at transition points between them (e.g. Fig. 2a, d, f and g).

Otolith Ba:Ca was also consistently low during the firstyear of life across all samples (Fig. 2). Minimum values(∼0.23–0.35 �mol mol−1) occurred prior to the first IP, with cycling

ries Re

i(tawpgca

2iwtfi4sncdmlstbo

4

tmsTaoaoaoda

fiacipdTBwsa1beOudv1mwm

J.I. Macdonald et al. / Fishe

n Ba:Ca evident in the outer otolith growth zones of most otolithse.g. Fig. 2a, d, e, f and g). The frequency, magnitude and shape ofhese peaks were highly variable. In some cases, cycles of Ba:Cand Sr:Ca appear to be in near synchrony and/or closely correlatedith the position of opaque growth zones, or to occur with annualeriodicity in at least some parts of the sample (e.g. Fig. 2a, d, f and). In others, Ba:Ca and Sr:Ca cycles appear de-coupled, with nolear relationship evident between cycles (or lack of) of Ba:Ca andnnulus deposition (e.g. Fig. 2b, c, e and h).

Concentrations of Mn:Ca typically reached maxima at between00 and 1000 �m from the otolith core, before gradually declin-

ng to minimum values near the otolith edge. Maximum valuesere sometimes associated with dramatic peaks at concentra-

ions up to ∼18 �mol mol−1 that coincided with the position of therst IP (Fig. 2d and h). More often, maximum values in the range–6 �mol mol−1 preceded the first IP (Fig. 2b, c, e and g) and occa-ionally occurred after it (Fig. 2a and f). Cu:Ca was slightly enrichedear the otolith core in some samples, with a small peak sometimesoncurrent with the initial rise in Mn:Ca (Fig. 2b). Mg:Ca valuesecreased slightly from maxima at, or within 200 �m of the core, toinimum values at varying points along the transects (usually fol-

owing formation of the 2nd annulus). Li:Ca data was variable, buthowed a consistent decreasing trend as the transect progressedowards the otolith edge. Pb:Ca values were uniformly low and sta-le across all transects but were always above the detection limitsf our system.

. Discussion

Accurate data on the spatial structuring of commercially impor-ant tuna stocks are crucial for defining appropriately scaled

anagement units that balance sustainable exploitation and con-ervation objectives (Fogarty and Botsford, 2007; Block et al., 2011).he results of this study indicate that trace element analysis oflbacore otoliths can potentially contribute valuable informationn the movement and mixing patterns of the South Pacific stockt a geographic scale useful for fishery management purposes. Byverlaying age-related information and sampling otolith materialt fine temporal resolution (i.e. <1 month of otolith growth at thetolith edge and ∼2 weeks at the otolith core), we were able toiscriminate capture locations of age 3+ and 4+ albacore with highccuracy (overall 85% of individuals correctly classified).

Additionally, although based on a very small dataset, the resultsrom the otolith core analyses provide some preliminary insightsnto connections between larval sources and adult populationscross the South Pacific Ocean. For example, the similarity in otolithore chemistry of New Zealand and New Caledonian fish spawnedn 2005/06 could imply that some mixing occurred between larvalools, or alternatively, that fish originated from geographicallyistant sources characterised by similar environmental conditions.he French Polynesian fish exhibited differences in Mg:Ca anda:Ca at the otolith core compared with the western capture sites,hich might suggest that they originated from a chemically and/or

patially distinct larval source. Our results provide no informations to the location of spawning sites or larval sources (see Ueyanagi,969; Yoshida, 1971; Nishikawa et al., 1985; Farley et al., 2013b)ut they do proffer questions on what degree of larval/juvenilexchange occurs between the eastern and western South Pacificcean, and do stock connectivity patterns remain consistentnder differing oceanographic/climatic conditions, for example,uring El Nino Southern Oscillation (ENSO)-neutral (i.e. 2005/06),ersus El Nino and La Nina years (Kimura et al., 1997; Lu et al.,

998; Briand et al., 2011)? Expanding the present analysis acrossultiple year-classes and regions, and integrating this informationith ocean-circulation hindcasts over the time series of interestay help address such questions.

search 148 (2013) 56– 63 61

Horizontal and vertical movement patterns of albacore areknown to be highly diverse, region-specific and follow seasonaltrends that appear closely associated with oceanographic con-ditions and prey availability (Otsu and Uchida, 1963; Laurs andLynn, 1977; Laurs et al., 1977; Chen et al., 2005; Domokos et al.,2007; Ichinokawa et al., 2008; Domokos, 2009; Sagarminaga andArrizabalaga, 2010; Briand et al., 2011; Childers et al., 2011). Of par-ticular relevance to the present study are the strong relationshipsevident between upwelling boundaries, oceanic fronts and eddies(which often display gradients in temperature, salinity, chlorophylla or micronekton) and albacore distribution (Domokos et al., 2007;Zainuddin et al., 2008). Several frontal zones exist in the westernand central South Pacific, with circulation of the East AustralianCurrent System (e.g. Tasman Front) and the convergence of warmsubtropical and cooler subantarctic waters (e.g. Subtropical Front)largely influencing their location and physical characteristics inlatitudes between New Zealand and New Caledonia (see Ridgwayand Dunn, 2003; Hamilton, 2006; Baird et al., 2008; Mullaney et al.,2011). Concentrations of Sr:Ca and Ba:Ca in otoliths can be sensi-tive to changes in salinity, temperature and diet (Bath et al., 2000;Elsdon and Gillanders, 2003; Macdonald and Crook, 2010; Waltheret al., 2010), and as precipitation of BaSO4 is elevated in areas ofhigh productivity (Stecher and Kogut, 1999), increases in otolithBa:Ca have been suggested to reflect periods of residence nearupwelling zones (Patterson et al., 2004; Kingsford et al., 2009; Wanget al., 2009). Given current knowledge of movement patterns inalbacore, it is conceivable that the Sr:Ca and Ba:Ca cycles that wedetected here and their close correspondence with the formationof opaque zones (fast summer growth) in some samples may reflectseasonal north-south movements across frontal zones (Langley,2004; Hoyle and Davies, 2009; Childers et al., 2011). Clear peri-odicity was not always evident however, indicating perhaps thatlarge-scale horizontal movements did not occur for those individ-uals, frontal gradients encountered were not strong during thoseyears, or that the resolution of our technique was not sufficientto detect environmental or developmental signals associated withmovement.

The discovery of enriched Mn concentrations at the otolith coresof several marine and freshwater fishes has improved precisionin the sampling of otolith material deposited within the egg, orat larval nursery sites (e.g. Brophy et al., 2004; Ruttenberg et al.,2005; Macdonald et al., 2008). Mn:Ca peaks were detected in almostall samples in the present study, but surprisingly, these did notcoincide with the otolith primordium. There was occasional cor-respondence between the Mn:Ca peak and the first IP, but peaksmore often occurred prior to its formation. Wang et al. (2009)found comparable results for southern bluefin tuna, and suggesteda link between the timing of Mn:Ca decrease and metamorpho-sis from larva to juvenile. It is possible that physiological and/ordietary changes associated with transition from juvenile to sub-adult stages may be driving the patterns in Mn:Ca we observed foralbacore (see de Pontual et al., 2003). Our findings could also reflectthe measurement of otolith material accreted later in life outsideof the target ‘otolith core’ region. Such measurement errors wouldmask Mn:Ca peaks at the core even if they were present. However,in light of the otolith preparation and analytical procedures used,we feel that this explanation is unlikely.

The high mobility of albacore dictates that rapid transitionsmust be made through differing ocean environments – a scenariowhich poses substantial challenges to disentangling the biologi-cal versus physical influences on elemental uptake into the otolith,and hence, for accurately reconstructing environmental histo-

ries. For example, a broad range of diurnal diving behaviours hasbeen noted for juvenile albacore in the North Pacific, with day-time depths exceeding 400 m for some individuals (Childers et al.,2011). Such dives exposed fish to large temperature fluctuations

6 ries Re

(aattBecltsdci�tfetmgl

A

DFWicLi

R

A

B

B

B

B

B

B

B

B

B

C

C

2 J.I. Macdonald et al. / Fishe

Childers et al., 2011), a factor that can strongly influence Srnd Ba uptake into otoliths (Elsdon and Gillanders, 2002; Martinnd Wuenschel, 2006). Moreover, recent indications of substan-ial time-lags (i.e. weeks to months) between a fish’s exposureo changed environmental conditions and when otolith Sr anda concentrations stabilise to reflect these new conditions (Lowet al., 2009; Macdonald and Crook, 2010; Miller, 2011) furtheromplicates data interpretation in wild-caught albacore, particu-arly given the lack of opportunities for controlled experiments onrace element uptake rates in this species. Nonetheless, the successeen here in discriminating capture locations suggests that furtherevelopment of these techniques for fishery management appli-ations is warranted. Other isotopic markers that reflect changesn metabolic processes and temperature conditions (i.e. �13C and18O) have proved effective for environmental history reconstruc-ion in fishes (Begg and Weidman, 2001; Weidel et al., 2007) andor delineating stock structure in tunas (Rooker et al., 2008; Shiaot al., 2009; Wells et al., 2012). Measuring these markers in conjunc-ion with trace elements, and indeed, combining inference from

ultiple sources (e.g. tagging experiments, commercial catch data,enetics, ocean circulation models) holds promise for resolvingongstanding questions on mixing and movement in albacore.

cknowledgements

This study was funded by the European Union 9th Europeanevelopment Fund through SCIFISH (Scientific Support for Oceanicisheries Management in the Western and Central Pacific Ocean).e extend thanks to the fisheries observers, crews and skippers

n the Pacific Island Countries and Territories and New Zealand forollecting samples. Thanks also to Alan Greig for assistance with theA-ICPMS analysis and Kaye Macdonald for comments that greatlymproved an earlier draft of the manuscript.

eferences

rrizabalaga, H., López-Rodas, V., Ortiz de Zárate, V., Costas, E., González-Garcés, A.,2002. Study on the migrations and stock structure of albacore (Thunnus alalunga)from the Atlantic Ocean and the Mediterranean Sea based on conventional tagrelease-recapture experiences. Collect. Vol. Sci. Pap. Int. Com. Conserv. AH. Tuna54, 1479–1494.

aird, M.E., Timko, P.G., Suthers, I.M., Middleton, J.H., Mullaney, T.J., Cox, D.R., 2008.Biological properties across the Tasman Front off southeast Australia. Deep-SeaRes. (Pt. I 55), 1438–1455.

ath, G.E., Thorrold, S.R., Jones, C.M., Campana, S.E., McLaren, J.W., Lam, J.W.H., 2000.Strontium and barium uptake in aragonitic otoliths of marine fish. Geochim.Cosmochim. Acta 64, 1705–1714.

egg, G.A., Weidman, C.R., 2001. Stable �13C and �18O isotopes in otoliths of haddockMelanogrammus aeglefinus from the northwest Atlantic Ocean. Mar. Ecol. Prog.Ser. 216, 223–233.

ertignac, M., Hampton, J., Coan, A.L., 1999. Estimates of exploitation rates for northPacific albacore, Thunnus alalunga, from tagging data. Fish. Bull. 97, 421–433.

ertignac, M., Lehodey, P., Hampton, J., 1996. An analysis of South Pacific albacoretagging data: estimation of movement patterns, growth and mortality rates. In:Working Paper No. 3, Sixth South Pacific Albacore Research Workshop, Raro-tonga, Cook Islands.

lock, B.A., Dewar, H., Blackwell, S.B., Williams, T.D., Prince, E.D., Farwell, C.J., Bous-tany, A., Teo, S.L.H., Seitz, A., Walli, A., Fudge, D., 2001. Migratory movements,depth preferences, and thermal biology of Atlantic bluefin tuna. Science 293,1310–1314.

lock, B.A., Jonsen, I.D., Jorgensen, S.J., Winship, A.J., Shaffer, S.A., Bograd, S.J., Hazen,E.L., Foley, D.G., Breed, G.A., Harrison, A-L., Ganong, J.E., Swithenbank, A., Castle-ton, M., Dewar, H., Mate, B.R., Shillinger, G.L., Schaefer, K.M., Benson, S.R., Weise,M.J., Henry, R.W., Costa, D.P., 2011. Tracking apex marine predator movementsin a dynamic ocean. Nature 475, 86–90.

riand, K., Molony, B., Lehody, P., 2011. A study on the variability of albacore (Thun-nus alalunga) longline catch rates in the southwest Pacific Ocean. Fish. Oceanogr20, 517–529.

rophy, D., Jeffries, T.E., Danilowicz, B.S., 2004. Elevated manganese concentrationsat the cores of clupeid otoliths: possible environmental, physiological, or struc-

tural origins. Mar. Biol. 144, 779–786.

ampana, S.E., 1999. Chemistry and composition of fish otoliths: pathways, mecha-nisms and applications. Mar. Ecol. Prog. Ser. 188, 263–297.

ampana, S.E., Nielson, J.D., 1985. Microstructure of fish otoliths. Can. J. Fish. Aquat.Sci. 42, 1014–1032.

search 148 (2013) 56– 63

Chen, I.C., Lee, P.F., Tzeng, W.N., 2005. Distribution of albacore (Thunnus alalunga) inthe Indian Ocean and its relation to environmental factors. Fish. Oceanogr. 14,71–80.

Childers, J., Snyder, S., Kohin, S., 2011. Migration and behavior of juvenile NorthPacific albacore (Thunnus alalunga). Fish. Oceanogr. 20, 157–173.

Dagorn, L., Bach, P., Josse, E., 2000. Movement patterns of large bigeye tuna (Thunnusobesus) in the open ocean determined using ultrasonic telemetry. Mar. Biol. 136,361–371.

Davies, C.A., Brophy, D., Jeffries, T., Gosling, E., 2011. Trace elements in the otolithsand dorsal spines of albacore tuna (Thunnus alalunga, Bonnaterre, 1788): anassessment of the effectiveness of cleaning procedures at removing postmortemcontamination. J. Exp. Mar. Biol. Ecol. 396, 162–170.

de Pontual, H., Lagardere, F., Amara, R., Bohn, M., Ogor, A., 2003. Influence of ontoge-netic and environmental changes in the otolith microchemistry of juvenile sole(Solea solea). J. Sea. Res. 50, 199–210.

Domokos, R., 2009. Environmental effects on forage and longline fishery per-formance for albacore (Thunnus alalunga) in the American Samoa ExclusiveEconomic Zone. Fish. Oceanogr. 18, 419–438.

Domokos, R., Seki, M.P., Polovina, J.J., Hawn, D.R., 2007. Oceanographic investigationof the American Samoa albacore (Thunnus alalunga) habitat and longline fishinggrounds. Fish. Oceanogr. 16, 555–572.

Eggins, S.M., Kinsley, L.P.J., Shelley, J.M.G., 1998. Deposition and element fraction-ation processes during atmospheric pressure laser sampling for analysis byICP–MS. Appl. Surf. Sci. 127–129, 278–286.

Elsdon, T.S., Gillanders, B.M., 2002. Interactive effects of temperature and salinity onotolith chemistry: challenges for determining environmental histories of fish.Can. J. Fish. Aquat. Sci. 59, 1796–1808.

Elsdon, T.S., Gillanders, B.M., 2003. Relationship between water and otolith elemen-tal concentrations in juvenile black bream Acanthopagrus butcheri. Mar. Ecol.Prog. Ser. 260, 263–272.

Farley, J.H., Williams, A.J., Clear, N.P., Davies, C.R., Nicol, S.J., 2013a. Age estima-tion and validation for South Pacific albacore Thunnus alalunga. J. Fish. Biol. 82,1523–1544.

Farley, J.H., Williams, A.J., Hoyle, S.D., Davies, C.R., Nicol, S.J., 2013b.Reproductive dynamics and potential annual fecundity of SouthPacific albacore tuna (Thunnus alalunga). PLoS ONE 8 (4), e60577,http://dx.doi.org/10.1371/journal.pone.0060577.

Fogarty, M.J., Botsford, L.W., 2007. Population connectivity and spatial managementof marine fisheries. Oceanography 20, 112–123.

García, A., Cortés, D., Ramírez, T., Fehri-Bedoui, R., Alemany, F., Rodríguez, J.M.,Carpena, A., Álvarez, J.P., 2006. First data on growth and nucleic acid and pro-tein content of field-captured Mediterranean bluefin (Thunnus thynnus) andalbacore (Thunnus alalunga) tuna larvae: a comparative study. Sci. Mar. 70S2,67–78.

Hamilton, L.J., 2006. Structure of the Subtropical Front in the Tasman Sea. Deep-SeaRes. (Pt. I 53), 1989–2009.

Hoover, R.R., Jones, C.M., 2013. Effect of laser ablation depth in otolith life historyscans. Mar. Ecol. Prog. Ser. 486, 247–256.

Hoyle, S., 2011. Stock assessment of albacore tuna in the South Pacific Ocean. In:Working paper SA-WP-06, 7th Scientific Committee meeting of the Western andCentral Pacific Fisheries Commission, Pohnpei, Federated States of Micronesia.

Hoyle, S., Davies, N., 2009. Stock assessment of albacore tuna on the South PacificOcean. In: WCPFC-SC5-2009 Working Paper No SA-WP-6. Port Vila, Vanuatu.

Ichinokawa, M., Coan, A.L., Takeuchi, Y., 2008. Transoceanic migration rates of youngNorth Pacific albacore, Thunnus alalunga, from conventional tagging data. Can.J. Fish. Aquat. Sci. 65, 1681–1691.

Kimura, S., Nakai, M., Sugimoto, T., 1997. Migration of albacore, Thunnus alalunga, inthe North Pacific Ocean in relation to large oceanic phenomena. Fish. Oceanogr.6, 51–57.

Kingsford, M.J., Hughes, J.M., Patterson, H.M., 2009. Otolith chemistry of the non-dispersing reef fish Acanthochromis polyacanthus: cross-shelf patterns from thecentral Great Barrier Reef. Mar. Ecol. Prog. Ser. 377, 279–288.

Kolody, D.S., Preece, A.L., Davies, C.R., Hartog, J.R., Dowling, N.A., 2010. Integratedevaluation of management strategies for tropical multi-species longline fish-eries. In: Final report for FRDC project 2007/017. CSIRO Marine and AtmosphericResearch, Hobart.

Labelle, M., 1993. A review of albacore tagging in the South Pacific. In: Tuna and Bill-fish Assessment Programme Technical Report No. 33, South Pacific Commission,Nouméa, New Caledonia.

Lan, K.W., Kawamura, H., Lee, M.A., Lu, H.J., Shimada, T., Hosoda, K., Sakaida, F.,2012. Relationship between albacore (Thunnus alalunga) fishing grounds in theIndian Ocean and the thermal environment revealed by cloud-free microwavesea surface temperature. Fish. Res. 113, 1–7.

Langley, A., 2004. An examination of the influence of recent oceanographic condi-tions on the catch rate of albacore in the main domestic longline fisheries. In:SCTB17 Working Paper No. SA-4. Nouméa, New Caledonia.

Laurs, R.M., Lynn, R.J., 1977. Seasonal migration of North Pacific albacore,Thunnus alalunga, into North American coastal waters: distribution, rela-tive abundance, and association with transition zone waters. Fish. Bull. 75,795–822.

Laurs, R.M., Nishimoto, R., Wetherall, J.A., 1985. Frequency of increment formation

on sagittae of North Pacific albacore (Thunnus alalunga). Can. J. Fish. Aquat. Sci.42, 1552–1555.

Laurs, R.M., Yuen, H., Johnson, J.H., 1977. Small-scale movements of albacore, Thun-nus alalunga, in relation to ocean features as indicated by ultrasonic tracking andoceanographic sampling. Fish. Bull. 75, 347–355.

ries Re

L

L

M

M

M

M

M

M

N

O

P

P

P

P

R

R

R

R

juvenile albacore, Thunnus alalunga, in the South Pacific Ocean. Fish. Bull. 69,

J.I. Macdonald et al. / Fishe

owe, M.R., DeVries, D.R., Wright, R.A., Ludsin, S.A., Fryer, B.J., 2009. Coastal large-mouth bass (Micropterus salmoides) movement in response to changing salinity.Can. J. Fish. Aquat. Sci. 66, 2174–2188.

u, H.J., Lee, K.T., Liao, H., 1998. On the relationship between El Nino Southernoscillation and South Pacific albacore. Fish. Res. 39, 1–7.

acdonald, J.I., Crook, D.A., 2010. Variability in Sr:Ca and Ba:Ca ratios in waterand fish otoliths across an estuarine salinity gradient. Mar. Ecol. Prog. Ser. 413,147–161.

acdonald, J.I., Shelley, J.M.G., Crook, D.A., 2008. A method for improving theestimation of natal chemical signatures in otoliths. Trans. Am. Fish. Soc. 137,1674–1682.

artin, G.B., Wuenschel, M.J., 2006. Effect of temperature and salinity on otolithelement incorporation in juvenile gray snapper Lutjanus griseus. Mar. Ecol. Prog.Ser. 324, 229–239.

ercier, L., Darnaude, A.M., Bruguier, O., Vasconcelos, R.P., Cabral, H.N., Costa, M.J.,Lara, M., Jones, D.L., Mouillot, D., 2011. Selecting statistical models and variablecombinations for optimal classification using otolith microchemistry. Ecol. Appl.21, 1352–1364.

iller, J.A., 2011. Effects of water temperature and barium concentration on otolithcomposition along a salinity gradient: implications for migratory reconstruct-ions. J. Exp. Mar. Biol. Ecol. 405, 42–52.

ullaney, T.J., Miskiewicz, A.G., Baird, M.E., Burns, P.T.P., Suthers, I.M., 2011. Entrain-ment of larval fish assemblages from the inner shelf into the East AustralianCurrent and into the western Tasman Front. Fish. Oceanogr. 20, 434–447.

ishikawa, Y., Honma, M., Ueyanagi, S., Kikawa, S., 1985. Average Distribution of Lar-vae of Oceanic Species of Scombroid Fishes, 1956–19. Far Seas Fisheries ResearchLaboratory, Shimizu, Japan, pp. 12.

tsu, T., Uchida, R.N., 1963. Model of the migration of albacore in the North PacificOcean. Fish. Bull. 63, 33–44.

aton, C., Hellstrom, J., Paul, B., Woodhead, J., Hergt, J., 2011. Iolite: freeware for thevisualisation and processing of mass spectrometer data. J. Anal. Atom. Spectrom.26, 2508–2518.

atterson, H.M., Kingsford, M.J., McCulloch, M.T., 2004. The influence of oceanic andlagoonal plume waters on otolith chemistry. Can. J. Fish. Aquat. Sci. 61, 898–904.

atterson, T.A., Evans, K., Carter, T.I., Gunn, J.S., 2008. Movement and behaviour oflarge southern bluefin tuna (Thunnus maccoyii) in the Australian region deter-mined using pop-up satellite archival tags. Fish. Oceanogr. 17, 352–367.

roctor, C.H., Thresher, R.E., 1998. Effects of specimen handling and otolith prepa-ration on concentration of elements in fish otoliths. Mar. Biol. 131, 681–694.

idgway, K.R., Dunn, J.R., 2003. Mesoscale structure of the mean East Australian Cur-rent System and its relationship with topography. Prog. Oceanogr. 56, 189–222.

ooker, J.R., Secor, D.H., Zdanowicz, V.S., Itoh, T., 2001. Discrimination of northernbluefin tuna from nursery areas in the Pacific Ocean using otolith chemistry.Mar. Ecol. Prog. Ser. 218, 275–282.

ooker, J.R., Secor, D.H., DeMetrio, G., Kaufman, A.J., Rios, A.B., Ticina, V., 2008. Evi-dence of trans-Atlantic movement and natal homing of bluefin tuna from stable

isotopes in otoliths. Mar. Ecol. Prog. Ser. 368, 231–239.

uttenberg, B.I., Hamilton, S.L., Hickford, M.J.H., Paradis, G.L., Sheehy, M.S., Standish,J.D., Ben-Tzvi, O., Warner, R.R., 2005. Elevated levels of trace elements in coresof otoliths and their potential for use as natural tags. Mar. Ecol. Prog. Ser. 297,273–281.

search 148 (2013) 56– 63 63

Sagarminaga, Y., Arrizabalaga, H., 2010. Spatio-temporal distribution of albacore(Thunnus alalunga) catches in the northeastern Atlantic: relationship with thethermal environment. Fish. Oceanogr. 19, 121–134.

Schloesser, R.W., Neilson, J.D., Secor, D.H., Rooker, J.R., 2010. Natal origin of Atlanticbluefin tuna (Thunnus thynnus) from Canadian waters based on otolith �13C and�18O. Can. J. Fish. Aquat. Sci. 67, 563–569.

Shiao, J.C., Yui, T.F., Høie, H., Ninnemann, U., Chang, S.K., 2009. Otolith O and C stableisotope composition of southern bluefin tuna Thunnus maccoyii (Pisces: Scom-bridae) as possible environmental and physiological indicators. Zool. Stud. 48,71–82.

Shiao, J.C., Wang, S.W., Yokawa, K., Ichinokawa, M., Takeuchi, Y., Chen, Y.G., Shen,C.C., 2010. Natal origin of Pacific bluefin tuna Thunnus orientalis inferred fromotolith oxygen isotope composition. Mar. Ecol. Prog. Ser. 420, 207–219.

Sibert, J., Hampton, J., 2003. Mobility of tropical tunas and the implications forfisheries management. Mar. Pol. 27, 87–95.

Stecher, H.A., Kogut, M.B., 1999. Rapid barium removal in the Delaware estuary.Geochim. Cosmochim. Acta 63, 1003–1012.

Swearer, S.E., Caselle, J.E., Lea, D.W., Warner, R.R., 1999. Larval retention and recruit-ment in an island population of a coral-reef fish. Nature 402, 799–802.

Ueyanagi, S., 1969. Observations on the distribution of tuna larvae in the Indo-Pacific Ocean with emphasis on the delineation of the spawning area of albacore,Thunnus alalunga. Bull. Far. Seas Fish. Res. Lab. 2, 177–219.

Walther, B.D., Kingsford, M.J., O’Callaghan, M.D., McCulloch, M.T., 2010. Interactiveeffects of ontogeny, food ration and temperature on elemental incorporation inotoliths of a coral reef fish. Environ. Biol. Fish. 89, 441–451.

Wang, C.H., Lin, Y.T., Shiao, J.C., You, C.F., Tzeng, W.N., 2009. Spatio-temporal varia-tion in the elemental compositions of otoliths of southern bluefin tuna Thunnusmaccoyii in the Indian Ocean and its ecological implication. J. Fish. Biol. 75,1173–1193.

Weidel, B.C., Ushikub, T., Carpenter, S.R., Kita, N.T., Cole, J.J., Kitchell, J.F., Pace,M.L., Valley, J.W., 2007. Diary of a bluegill (Lepomis marcrochirus): daily �13Cand �18O records in otoliths by ion microprobe. Can. J. Fish. Aquat. Sci. 64,1641–1645.

Wells, R.J.D., Rooker, J.R., Itano, D.G., 2012. Nursery origin of yellowfin tuna in theHawaiian Islands. Mar. Ecol. Prog. Ser. 461, 187–196.

White, J.W., Ruttenberg, B.I., 2007. Discriminant function analysis in marine ecol-ogy: some common oversights and their solutions. Mar. Ecol. Prog. Ser. 329,301–305.

Williams, A.J., Nicol, S.N., Leroy, B., 2010. South Pacific albacore tagging project: 2010summary report. In: Report for the Western and Central Fisheries Commission,Scientific Committee Sixth Regular Session, Nuku’alofa, Tonga.

Woodhead, J., Swearer, S., Hergt, J., Maas, R., 2005. In situ Sr-isotope analysis ofcarbonates by LA-MC-ICP-MS: interference corrections, high spatial resolutionand an example from otolith studies. J. Anal. At. Spectrom. 20, 22–27.

Yoshida, H., 1971. Distribution, apparent abundance, and length composition of

821–827.Zainuddin, M., Saitoh, K., Saitoh, S., 2008. Albacore (Thunnus alalunga) fishing ground

in relation to oceanographic conditions in the western North Pacific Ocean usingremotely sensed satellite data. Fish. Oceanogr. 17, 61–73.