Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

Contents lists available at ScienceDirect

Journal of Experimental Marine Biology and Ecology

j ourna l homepage: www.e lsev ie r .com/ locate / jembe

Inferringmovement patterns of a coral reef fish using oxygen and carbonisotopes in otolith carbonate

Leanne M. Currey a,b,c,⁎, Michelle R. Heupel b,c, Colin A. Simpfendorfer c, Ashley J. Williams c,d

a AIMS@JCU, James Cook University, Townsville, Queensland 4811, Australiab Australian Institute of Marine Science, PMB No 3, Townsville, 4810, Australiac Centre for Sustainable Tropical Fisheries and Aquaculture, School of Earth and Environmental Sciences, James Cook University, Townsville, Queensland 4811, Australiad Oceanic Fisheries Programme, Secretariat of the Pacific Community, BP D5, 98848 Noumea, New Caledonia

⁎ Corresponding author at: AIMS@JCU, James Cook Uni4811, Australia. Tel.: +61 7 4781 4158.

E-mail addresses: leanne.currey@my.jcu.edu.au (L.M. C(M.R. Heupel), colin.simpfendorfer@jcu.edu.au (C.A. Simp(A.J. Williams).

http://dx.doi.org/10.1016/j.jembe.2014.03.0040022-0981/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 11 October 2013Received in revised form 31 January 2014Accepted 9 March 2014Available online xxxx

Keywords:Broad-scale movementLethrinus miniatusMigrationOntogenyStable isotope chemistry

Ratios of oxygen and carbon isotopes (δ18O and δ13C) in otolith carbonate provide information on stockseparation and migration of fish populations inhabiting isotopically different environments. δ18O and δ13Cwere analysed from Lethrinus miniatus (Forster 1801) otoliths to elucidate the probability of broad-scalemovement in the Great Barrier Reef. Otolith core and edge portions were compared between and withinindividuals from the same cohort (n = 120) to determine whether isotopically different environments wereinhabited by juveniles and adults from different latitudes. A comparison between individuals revealed thatδ18O and δ13C increased with latitude and were generally higher in edge than in core portions. Similarly, withinmost individual otoliths, δ18O and δ13C were higher in edge than in core portions. For some individuals, however,there was no difference in δ18O between core and edge portions, potentially signifying self-recruitment to thearea or occupation of areas of similar chemistry. The majority of individuals from 19 to 22°S and almost half ofindividuals from 18°S displayed juvenile δ18O signatures characteristic of a warmer environment, suggestingmovement to cooler environments or potential southerly migration of early life stages to adult habitat. Incontrast, almost half of adult individuals in the most northern latitude (18°S) appeared to originate from coolerenvironments, suggestingmovement towarmer environments or potential northwardmigrationwith ontogeny.These results infer thatmovement of L. miniatus individuals occurred across different isotopic environmentswithlife stage and was related to latitude. Isotope ratios provided insight into movement patterns of L. miniatus atscales that will be important for delineating management units for this commercially important species.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Movement is a fundamental demographic variable affecting thedistribution and abundance of populations in time and space (Pittmanand McAlpine, 2001). Knowledge of movement patterns in fishes isimportant for understanding population dynamics, delineating stockboundaries and evaluating the potential effects of exploitation (Cadrinand Secor, 2009; Fromentin and Powers, 2005). Movement patterns ofexploited coral reef fishes have not been extensively studied, despitethe strong influence of movement on the distribution and abundanceof fishes (Pittman and McAlpine, 2001). Coral reef fish larvae aretypically pelagic and have the potential to disperse widely (Cowen,1990), while after settlement, most are considered sedentary withrelatively small home ranges (Sale, 1991). This paradigm evolved

versity, Townsville, Queensland

urrey), m.heupel@aims.gov.aufendorfer), ashleyw@spc.int

principally from studies of small-bodied reef fish and may not apply tosome larger-bodied species that have the capacity to move largerdistances. Long-distance movements by a proportion of the adultpopulation (Grüss et al., 2011) can assist in connectivity of populations(Kaunda-Arara and Rose, 2004a). Thus, information on the dispersal ofboth the larval and adult stages is important for a more completeunderstanding of the population dynamics of mobile species.

Dispersal can be investigated by analysis of fish otoliths, which arecalcified aragonite structures that assist with balance and hearing.Otoliths act as time recorders, incorporating the elemental and isotopicsignatures of the environments a fish has inhabited as it grows(Campana, 1999, 2005). For a number of exploited species, otolithshave been used to identify migrations and discern stocks in isotopicallydistinct environments (Bastow et al., 2002; Edmonds and Fletcher,1997; Fairclough et al., 2013; Newman et al., 2009, 2011). Analysis ofstock structure typically involves obtaining a combined stable isotopicsignature for the whole otolith (e.g. Newman et al., 2010), while timingof migrations is investigated from portions of the otolith correspondingto growth structures (e.g. Steer et al., 2010). Ratios of stable oxygenisotopes (18O/16O or δ18O) are incorporated into otoliths at or near

19L.M. Currey et al. / Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

equilibriumwith environmental δ18O, varying as a function of tempera-ture and salinity (Elsdon and Gillanders, 2002), where a decrease inδ18O correlates to an increase in temperature and decreased salinity(Kalish, 1991). The ratio of stable carbon isotopes (13C/12C or δ13C) isnot incorporated into otoliths at equilibrium and is influenced bymeta-bolic rates and reflects diet and geographic variations in dissolved in-organic carbon (DIC) (Campana, 1999; Elsdon et al., 2010). Highmetabolism as juveniles results in lower δ13C than for adult fisheswhose metabolic rate has slowed (Høie et al., 2004). Investigation ofδ18O and δ13C at both juvenile and adult life stages can elucidate stockmixing or separation, movement or residency among locations andcan allow adults to be linked to nursery areas (Gillanders et al., 2003;Thorrold et al., 1998).

Although research on isotope analysis of otoliths has concentratedon temperate fishes (but see: Dufour et al., 1998), this techniquecould be applied to better understand the dispersal of tropical reef fishessuch as emperors (family Lethrinidae). Emperors are small to medium-bodied coral reef fishes that are often important components of tropicalsubsistence, artisanal and commercial fisheries throughout the Indo-West Pacific and West Africa (Carpenter, 2001). In comparison tomost other reef fish families, emperors are generally assumed to bemore mobile (Jones, 1991) because they are often encountered awayfrom reefs in sand or rubble areas, and usually do not appear to beterritorial (Carpenter, 2001; Kaunda-Arara and Rose, 2004a, 2004b).However, there is limited empirical movement data available for em-perors (but see: Nanami and Yamada, 2009; Taylor and Mills, 2013),and the majority of tagged individuals have been recaptured within5 km of release sites, with only a few individuals moving greater dis-tances (Kaunda-Arara and Rose, 2004a, 2004b). It remains unclear,therefore, whether emperors are more mobile than other reef fishesand whether regular large-scale dispersal is characteristic.

The redthroat emperor Lethrinus miniatus is one of the largestemperors, attaining a length in excess of 600 mm fork length (FL) anda weight of over 3 kg (Williams et al., 2003, 2007a). It is an importantemperor species in commercial and recreational hook and line fisheriesin Australia, Japan, New Caledonia and Tonga (Carpenter, 2001), withthe largest fishery located in the Great Barrier Reef (GBR) (Leigh et al.,2006). Despite its importance to fisheries, there is limited empiricalevidence of adult movement of L. miniatus, due to low recapture ratesfrom tagging (Brown et al., 1994). For example, only 32 of 1500 exter-nally tagged individual L. miniatuswere recaptured over a 20 year peri-od on the GBR (B. Sawynock unpublished results;Williams et al., 2010).Biological research has reported significant regional differences in agestructure and natural mortality rates of L. miniatus (Williams et al.,2007b) and incorporating movement of L. miniatus among regions inan age-structured population model provided a parsimonious explana-tion for these observed differences (Williams et al., 2010). Williamset al. (2010) hypothesised that net movement of L. miniatus occurs ina northerly direction, away from the centre of the species' distributionon the GBR, or in a cross-shelf direction. However, there is no empiricalevidence to support any hypotheses on movement, and it is unlikelythat conventional tagging approaches can provide such evidence. Alter-native techniques such as the analysis of isotopic signatures in otolithsprovide the potential to gain information on movement patterns ofL. miniatus.

The aim of this study was to determine whether L. miniatus under-take broad-scale migration with ontogeny by: (1) identifying whetherotolith isotopic signatures of L. miniatus vary between life stage(juvenile and adult) with latitude; (2) comparing isotopic signaturesby latitude for individual otoliths; and (3) using these findings tospeculate on movement of this species. If isotopically different waterbodies were inhabited by juveniles and adults, isotopic signatures ofthe otoliths will reflect these differences. Using this novel approach,detection of variation in isotopic signatures with latitude and life stagemay explain regional differences in population demographics. Thiswill therefore provide insight into whether the population undergoes

directed long-distancemovementwith ontogeny, advancing our under-standing of movement patterns and providing valuable input for con-sideration in future stock assessments for L. miniatus.

2. Methods

2.1. Sample collection and processing

Samples of L. miniatuswere collected from reefs of the Great BarrierReef (GBR; Fig. 1) from research line fishing catch surveys for the Effectsof Line Fishing (ELF) experiment (Mapstone et al., 2004) and FisheriesResearch and Development Corporation (FRDC) Project No. 98/131(Davies et al., 2006). For each fish, FLwas recorded to the nearest mm,and sagittal otoliths were removed, cleaned, dried, weighed and storedin paper envelopes.

Otoliths were sectioned with a diamond blade isomet saw andmounted on glass slides with clear-casting polyester resin. Pairs oftranslucent and opaque bands have been previously validated as annuli(Williams et al., 2005) and age estimates of individuals were verified bycounts of completely formed opaque bands under reflected light. Fouryear old individuals were selected for analyses to minimise temporalvariability. September 1 was assigned as the birth date of this species,since this represents the middle of the spawning period (July–October;Williams et al., 2006). One hundred and twenty individual four year oldfish caught between 01/09/1999 and 31/08/2000 were selected foranalysis across latitudes 18°S to 22°S (Fig. 1).

2.2. Laboratory processing and analysis of isotope ratios

Slides of sectioned otoliths were cleaned usingmethanol and placedon a moveable microscope stage fitted with a digital camera and posi-tioned beneath a fixed high-precision micromilling bit. Two samplesfrom each otolith were selected for micromilling; the core (primordi-um) portion represented the first ~3 months of life and the edge por-tion referred to the final 3–6 months of life prior to capture relative tovisible otolith growth structures (Fig. 2). Sample paths were definedand digitised as three-dimensional coordinates using New WaveResearch XPC™ software (ESI Inc., Portland, OR, USA) (Hanson et al.,2010; Wurster et al., 1999). A predetermined amount of material wasselected to obtain 50–120 μm material for each portion, involvingthree paths at the core and a singular path along the otolith edge(Fig. 2). Samples were micromilled using the edge of a micro dentaldrill (0.3 mm) set to a width of 60 μm and depth of 120–140 μm.Resulting sample powder was collected manually using a scalpel tipand stored in small plastic 0.2ml PCR Eppendorf tubes prior toweighingand analysis.

δ18O and δ13C values were determined using a ThermoFisherGasBench III coupled to a Delta VPlus isotope ratio mass spectrometervia a Conflo IV. Sample powders were placed in glass exetainer vialsand atmosphere was replaced with helium. Complete digestion ofsample powders with 100% anhydrous orthophosphoric acid producedcarbon dioxide and water. Isotope ratios were analysed relative torepeat samples of international reference standards (NBS-19 limestoneand NBS-18 calcite) to monitor accuracy and precision. Precision ofinternal standards (standard deviation, SD, of mean values) wereb0.1‰ for both oxygen and carbon isotopes. Samples are reportedrelative to Vienna Pee Dee Belemnite (VPDB).

2.3. Statistical analyses

Variation in isotopic signatures with latitude indicates differences inenvironmental chemistry with location, while dissimilarity in δ18O andδ13C between life stages suggests individuals inhabited isotopicallydifferent environments during the juvenile and adult stages. To deter-mine whether differences in isotopic signature (δ18O and δ13C) existedbetween life stage (otolith portion) and latitude, linear regression

Bowen

Mackay

Cairns

Townsville

150°145°

17°

18°

19°

20°

21°

22°

23°

0 75 150

km

Australia

QUEENSLAND

Coral Sea

0 1,000 2,000

km

Fig. 1. Locations (black circles) where L. miniatus were sampled from the Great Barrier Reef World Heritage Area, Queensland Australia.

20 L.M. Currey et al. / Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

(δ18O) and generalised least squares models (GLS; δ13C) with an infor-mation theoretic model selection process in R version 3.0 (R Develop-ment Core Team: www.r-project.org) were used. GLS models areweighted linear regressions, inwhich heterogeneity of residual variancecan be incorporated into the analysis (Pinheiro et al., 2013; Zuur et al.,2009). Residual spread increased with portion for δ13C, so the varIdentvariance structure was used in the nlme package (Pinheiro et al.,2013) to weight the models by portion to achieve homogeneousvariances (Zuur et al., 2009). δ18O and δ13C were treated as responsevariables in separate analyses with latitude as a continuous factor andportion modelled as a fixed factor, month (capture date) as a naturalspline (with varying degrees of freedom) and whole otolith weight asan offset. Otolith weight is commonly included as a covariate in studiesinvestigating variation in isotope ratios that include individuals com-prising multiple age classes (e.g. Bastow et al., 2002; Newman et al.,

Fig. 2. Sectioned L. miniatus otolith under reflected light with user-de

2010). Although all fish in this study were collected from the samecohort, whole otolith weight was included in the models to minimisevariability (between birth and capture date) in the analysis. To deter-mine the best fitting model for both isotope ratios, models werecompared using the Akaike Information Criterion corrected for smallsample bias (AICc). Twenty-one models were compared for oxygenand carbon isotope ratios. Models were compared using a stepwiseapproach, from the most complex fully saturated model to the simplestmodel. Strongly supported models were those with ΔAICc values b2.Where multiple models satisfied this criterion, the best-approximatingmodel was considered as the model with the lowest ΔAICc and thefewest terms.

To examine the variation in δ18O and δ13C between life stages foreach individual otolith, the difference (d) between edge and core valuesfor each individual otolith was compared, where d= IE − IC, and E was

fined micromilling paths highlighted as core and edge portions.

Table 1Linear regression and GLSmodels examining the effects of latitude (L), otolith portion (P),whole otolith weight (W) and month (M), on δ18O and δ13C. Onlymodels with at least 1%support (Akaike weight (w) = 0.01) are shown. The effect of month was modelled as anatural spline (ns) with varying degrees of freedom. Otolith weight was modelled as anoffset. All δ13C GLSmodelswereweighted by portion using the varIdent variance structure.AICc is the small-sample bias-corrected formofAkaike's information criterion,ΔAICc is theAkaike difference. Models with ΔAICc b2 (in bold) are the best fitting models. Eachmodelwas compared to the null model, with significance given as p.

# Model AICc ΔAICc w p

O1 lm (δ18O ~ L × P + ns(M, df = 3) + offset (W))

20.85 0 0.23 b0.001

O2 lm (δ18O ~ L × P + ns (M, df = 3)) 21.32 0.47 0.19 b0.001O3 lm (δ18O ~ L × P + ns

(M, df = 4) + offset (W))21.90 1.05 0.14 b0.001

O4 lm (δ18O ~ L × P + ns (M, df = 4)) 22.65 1.81 0.10 b0.001O5 lm (δ18O ~ L × P + offset(W)) 23.26 2.42 0.07 b0.001O6 lm (δ18O ~ L × P + ns

(M, df = 2) + offset(W))23.32 2.47 0.07 b0.001

O7 lm (δ18O ~ L × P + ns (M, df = 2)) 23.73 2.89 0.06 b0.001O8 lm (δ18O ~ L × P) 24.40 3.56 0.04 b0.001O9 lm (δ18O ~ L + P + ns

(M, df = 3) + offset (W))25.07 4.23 0.03 b0.001

O10 lm (δ18O ~ L + P + ns (M, df = 3)) 25.42 4.58 0.02 b0.001O11 lm (δ18O ~ L + P + ns

(M, df = 4) + offset (W))26.15 5.30 0.02 b0.001

O12 lm (δ18O ~ L + P + ns (M, df = 4)) 26.77 5.93 0.01 b0.001O13 lm (δ18O ~ L + P + offset (W)) 27.12 6.27 0.01 b0.001C1 gls (δ13C ~ L × P + offset (W)) 725.16 0 0.28 b0.001C2 gls (δ13C ~ L × P) 725.16 0 0.28 b0.001C3 gls (δ13C ~ L × P + ns

(M, df = 2) + offset (W))727.98 2.82 0.07 b0.001

C4 gls (δ13C ~ L × P + ns (M, df = 2)) 727.98 2.82 0.07 b0.001C5 gls (δ13C ~ L × P + ns

(M, df = 3) + offset (W))728.00 2.84 0.07 b0.001

C6 gls (δ13C ~ L × P + ns (M, df = 3)) 728.00 2.84 0.07 b0.001C7 gls (δ13C ~ L + P + offset (W)) 729.11 3.95 0.04 b0.001C8 gls (δ13C ~ L + P) 729.11 3.95 0.04 b0.001C9 gls (δ13C ~ L × P + ns

(M, df = 4) + offset (W))730.55 5.39 0.02 b0.001

C10 gls (δ13C ~ L × P + ns (M, df = 4)) 730.55 5.39 0.02 b0.001

Fig. 3.Predicted trends in oxygen isotope ratios (δ18O) for (A) the interaction of core (opencircles; dashed line) and edge (closed circles; solid line) portions by latitude for themonthof September, and (B) month (1–12; January to December) at latitude 19.21°S. Linesindicate the fit of best-approximating model (O1) and 95% confidence intervals (greyshading) are shown.

21L.M. Currey et al. / Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

the edge value and C was the core value for each isotope I. Thisstandardization allowed changes in isotopic ratios with ontogeny tobe illustrated. Plots were produced with C normalised to zero, therefored N 0 represents an increase and d b 0 represents a decrease in δ18O andδ13Cwith life stage. If no difference in δ18O and δ13C existed between lifestages (i.e. d= 0), it would be assumed that the environments an indi-vidual inhabited as a juvenile and adultwere similar inwater chemistry,or no movement (among latitudes) occurred with ontogeny. Linear re-gressionmodelswere used to analyse δ18O and δ13C (14models for eachisotope) and were fitted with a natural spline of month, otolith weightas an offset, and compared using AICc.

The visreg package in R (Breheny and Burchett, 2013) was used tovisualise the regression models, to describe how the isotope ratiosvary as a function of the explanatory variables. Conditional plots foreach explanatory variable with each isotope ratio were producedwhile holding all other variables constant (as the median or mostcommon category), for example: δ18O by latitude for the month ofSeptember (Breheny and Burchett, 2013).

3. Results

Four models for δ18O were strongly supported by the data (ΔAICc b2) (models O1–O4, Table 1) and included parameters for latitude,otolith portion, the interaction between them, and month (Table 1).Each of these models were very similar with the same main effects,however, the best-approximating model (model O1) indicated thatδ18O increased with latitude. This trend was stronger for edge (adult)portions than for the core (juvenile) portions despite some variationobserved with latitude (Fig. 3A). δ18O varied among months and was

greater in samples collected during May/June and December, andlowest in January and September/October (Fig. 3B).

Two GLS models were unambiguously the best-approximatingmodels for δ13C (models C1 and C2, Table 1). Both models includedparameters for latitude, otolith portion and the interaction betweenthem. The simplest model (model C2) predicted an increase in δ13Cwith latitude, and higher δ13C for adult otolith portions than for juvenileotolith portions (Fig. 4).

For the analysis of life stage differences in isotope ratios withinindividuals, the best-approximating model for dδ18O included para-meters for latitude and month (model OA, Table 2). There was similarsupport (ΔAICc b 2) for a model with a higher degree natural splineon month (model OC) and model OB (Table 2). dδ18O increased withlatitude (Fig. 5A) and decreased in samples collected from January toAugust (Fig. 5B). Individual otoliths with dδ18O = 0 (dashed line;Fig. 5A and B) signify no difference in δ18O from core to edge, dδ18O b

0 signify δ18O decreased from core to edge, and dδ18 N 0 signify δ18O

Fig. 4. Predicted trends in carbon isotope ratios (δ13C) for the interaction of core (opencircles; dashed line) and edge (closed circles; solid line) portions by latitude. Lines indicatethe fit of best-approximatingmodel (C2) and 95% confidence intervals (grey shading) areshown.

22 L.M. Currey et al. / Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

increased from core to edge (Fig. 5A: data points). No difference in δ18Owas observed for fifteen individuals (10–13.3%, among latitudes). δ18Oincreased with life stage for the majority of individual otoliths overall(74%), particularly for those collected from 19 to 22°S (71.1%). Of theindividuals collected between latitudes 18 and 19°S, however, 43.3%decreased in δ18O and 43.3% increased in δ18O with life stage (Fig. 5A:data points).

Two linear regression models for dδ13C were strongly supported bythe data (ΔAICc b 2) and included parameters for latitude and month(models CA and CB, Table 2). The best-approximating model (CA)predicted dδ13C increased with latitude (Fig. 6A) and a greater dδ13C

Table 2Linear regression models examine the effects of latitude (L), whole otolith weight (W),month (M), on the difference (d) between edge and core δ18O values for each individualotolith. Only models with at least 1% support (Akaike weight (w) = 0.01) are shown.The effect of month was modelled as a natural spline (ns) with varying degrees offreedom. Otolith weight was modelled as an offset. AICc is the small-sample bias-corrected form of Akaike's information criterion, ΔAICc is the Akaike difference. Modelswith ΔAICc b2 (in bold) are the best fitting models. Each model was compared to thenull model, with significance given as p.

# Model AICc ΔAICc w p

OA lm (dδ18O ~ L + ns (M, df = 2)) 48.99 0 0.35 b0.001OB lm (dδ18O ~ L + ns

(M, df = 2) + offset (W))50.11 1.12 0.20 b0.001

OC lm (dδ18O ~ L + ns (M, df = 3)) 50.78 1.79 0.14 b0.001OD lm (dδ18O ~ L) 51.62 2.63 0.09 b0.001OE lm (dδ18O ~ L + ns

(M, df = 3) + offset (W))51.91 2.92 0.08 b0.001

OF lm (dδ18O ~ L + offset (W)) 52.49 3.50 0.06 b0.001OG lm (dδ18O ~ L + ns (M, df = 4)) 53.04 4.05 0.05 b0.001OH lm (dδ18O ~ L + ns

(M, df = 4) + offset (W))54.17 5.18 0.03 b0.001

CA lm (dδ13C ~ L + ns(M, df = 2) + offset (W))

414.23 0 0.27 b0.001

CB lm (dδ13C ~ L + ns (M, df = 2)) 414.40 0.17 0.25 b0.001CC lm (dδ13C ~ L + offset (W)) 416.50 2.27 0.09 b0.001CD lm (dδ13C ~ L + ns

(M, df = 3) + offset (W))416.56 2.33 0.09 0.002

CE lm (dδ13C ~ L) 416.64 2.41 0.08 b0.001CF lm (dδ13C ~ L + ns (M, df = 3)) 416.73 2.50 0.08 0.002CG lm (dδ13C ~ L + ns

(M, df = 4) + offset (W))417.34 3.11 0.06 0.002

CH lm (dδ13C ~ L + ns (M, df = 4)) 417.46 3.23 0.05 0.002

Fig. 5. Predicted trends in the difference in oxygen isotope ratios (dδ18O; points) betweenedge and core portions by individual, illustrating the relationship of best-approximatingmodel (OA; lines) with (A) latitude for the month of September, and (B) month (1–12;January to December) at latitude 19.21°S. Grey shading indicates 95% confidence intervalsand dashed line represents zero dδ18O.

during the months of June to September (Fig. 6B). As dδ13C N 0 for allindividuals, δ13C increased for all individual otoliths from core to edgeand was greater at higher latitudes. However, greater individualvariation in dδ13C was observed at higher latitudes, with lower, moreconsistent values of δ13C among northern individuals compared toindividuals from the south (Fig. 6A: data points).

4. Discussion

Results of the isotope analysis for L. miniatus otoliths providedevidence that isotope ratios of both oxygen (generally related to watertemperature) and carbon (which reflects DIC, diet and metabolism)increased with latitude and life stage. Specifically, δ18O and δ13Cincreased with latitude for edge (adult) portions, were similar (δ18O)or slightly increased (δ13C) with latitude for core (juvenile) portions,and the difference in isotope ratios with ontogeny increased with lati-tude within individuals. These findings suggest the potential for

Fig. 6. Predicted trends in the difference in carbon isotope ratios (dδ13C; points) betweenedge and core portions by individual, illustrating the relationship of best-approximatingmodel (CA; lines)with (A) latitude for themonth of September, (B)month (1–12; Januaryto December) at latitude 19.21°S. Grey shading indicates 95% confidence intervals.

23L.M. Currey et al. / Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

broad-scale ontogenetic movement of L. miniatus through areas of dif-ferent chemistry which may explain regional biological differences. Re-sults of this study indicate that isotope ratios formost individuals variedwith ontogeny, and the increase in isotope ratios with latitude suggeststhatmovement over a period of years is likely for this species, potential-ly at a scale of 10s of kilometres.

At the outset, it was important to determine whether isotopicallydifferent environments were present (and detectable) at the latitudesfromwhich sampleswere collected. Less variation inwater temperatureand salinity may exist in tropical regions compared to temperatesystems that can vary N9 °C along a salinity gradient (e.g. Edmondsand Fletcher, 1997; Fairclough et al., 2013; Steer et al., 2010). However,a trend for increased δ18O and δ13C for otolith edge portions withlatitude indicated different capture locations despite variability amongsamples. Variation in depth use could have potentially confounded theisotope ratios due to environmental variations in temperature, salinityand DIC. However, the overall effect of depth on δ18O and δ13C in

L. miniatus otoliths in the coral reef environment was thought to beminimal, since δ18O is similar down to 500 m depth due to verticalmixing (Craig and Gordon, 1965). Thus, this method was sensitiveenough to differentiate subtle trends in isotope ratios with latitude.

Oxygen isotope ratios in otolith portions revealed that signaturesamong adults were different, but similar among juveniles with latitude.δ18O generally increased for adult edge portions with latitude,which reflected the difference in capture location (18–22°S) and wasconsistent with higher values of δ18O at cooler water temperatures(Kalish, 1991). The juvenile habitat remains unknown and can only beestimated as there is currently no information available about larval orsettlement phases of L. miniatus and no juveniles (b20 cm FL) havebeen observed. The δ18O signatures in otolith cores likely incorporatedthe pelagic larval phase through to settlement and early juvenilephase. Thus, similarities among core portions with latitude indicatedthat juveniles originated from a similar location, or occupied locationswith similar oceanography within the first fewmonths of life. Althoughtemporal variationwasminimised in this study (i.e. individuals selectedfrom the same cohort), some influence of month of capture and otolithweight was observed. The absence of clear patterns in δ18O with monthof capture and otolithweight was likely due to variation among individ-uals, rather than any temporal change in the environment (e.g. seasonalwater temperature).

Analysing the difference between the juvenile (core) and adult(edge) δ18O within individuals provided a greater understanding ofontogenetic movement of L. miniatus between different isotopic envi-ronments. Although multiple samples across an otolith transect wouldprovide higher-resolution information with ontogeny, the two samplesper individual obtained in this study provide the first insight into usingisotopes to investigate broad-scale movement for L. miniatus. As coreδ18O values were similar among latitudes, a change in δ18O with on-togeny could indicate movement to the adult location. A small pro-portion of individuals from each latitude displayed no difference inδ18O with ontogeny, implying self-recruitment by a resident proportionof the population (Grüss et al., 2011), or that these individuals inhabitedenvironments with similar chemistry as juveniles and adults. Withoutsimultaneous isotopic measurements from capture locations, we canonly infer possible explanations for the overall pattern of increasedδ18O with latitude from the juvenile to adult stage. Yet, differenceswith ontogeny signify movement between water bodies, and individ-uals captured from southern latitudes indicated that juveniles inhabitedwarmer environments compared to the capture locations of adults (i.e.overall increase in δ18O with ontogeny). This suggests that adultL. miniatus from cooler waters in the southern GBRmay have originatedfrom warmer waters with potentially higher salinity in the north, orthey originated from juveniles that resided in warmer shallow habitats,or nearby environments with similar chemistry. The first suggestion isconsistent with larval dispersal from north to south, in the direction ofthe Eastern Australian Current (Bode et al., 2006). Latitudinal patternsin δ18O also partly supportWilliams et al.'s (2010) hypothesis formigra-tion of post-recruits away from the middle of the GBR, with southerlymovement of individuals from ~20.5°S. It is unlikely that juvenilesreside in warmer shallow habitats because no juveniles (b20 cm FL)have been observed despite many decades of research focussed inthese depths. Movement of individuals through proximate locationsof distinct water chemistry would also produce different isotopicsignatures with ontogeny, implying that movement occurs at a smallerspatial scale. The overall pattern with latitude, however, suggests thatthe potential for broader-scale movement is more likely.

In comparison, almost half of the far north individuals (~latitude18°S) showed a large decrease in δ18O with ontogeny, signifying theseindividuals occupied cooler habitats as juveniles. Northern migrationmay be one potential explanation for this finding (Williams et al.,2010). The concept that northern GBR reefs act as sources of larvae fordispersal to southern reefs has been described for other taxa (Bodeet al., 2006; Doherty, 2006) and was suggested for L. miniatus due to a

24 L.M. Currey et al. / Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

greater proportion of spawning females occurring in waters north of20°S compared to southern waters (Williams et al., 2006). In contrast,the increase in δ18Owith ontogeny for almost half of the adults capturedat northern latitudes implies that two different strategies may beemployed, but this is not entirely surprising. Not all individuals from apopulation conform to exactly the same movement pattern; somemay be residents, while others may be more wide-ranging individuals(e.g. Kaunda-Arara and Rose, 2004a; Moran et al., 2003) which can beintegral to population connectivity. Movement of L. miniatus to latitude18°S from the north is possible since L. miniatus do occur infrequently to16°S (Williams and Russ, 1994). Adult L. miniatus are also thought tomove in response to extreme weather events. Tobin et al. (2010)suggested that increased catch rates in northern latitudes after cyclones(e.g. tropical cyclone Justin) was due to adults migrating to shallowreefs from deep shelf and shoal waters, with the advection of deepcold water onto the continental shelf. Alternatively, if different chemis-tries were exhibited by nearby habitats which would result in differ-ences in otolith δ18O, movement nevertheless occurs with ontogeny.The trend for lower dδ18O observed for individuals collected later inthe year (capture month) did not provide any insight into potentialmovement patterns. Thus, further sample collection from reefs in thenorthern region would provide more clarity.

Across all latitudes, carbon isotope ratios were greater for adult edgeportions compared to juvenile core portions, which were significantlylower (and less varied) in δ13C. Lower δ13C in the juvenile stage reflectsa different diet and a higher metabolic rate than for adults (Elsdon andGillanders, 2002; Høie et al., 2003; Thorrold et al., 1997) and is charac-teristic of temperate (Ayvazian et al., 2004; Schwarcz et al., 1998;Weidman andMillner, 2000) and tropical fishes, including other speciesof the genus Lethrinus (Huxam et al., 2007). While carbon isotopes arenot directly influenced by temperature, cyclic patterns in δ13C thatroughly followed patterns of δ18O have been attributed to metabolism(Høie et al., 2004) because seasonal shifts in diet were controlled for(in contrast to Weidman and Millner, 2000). This temperature-induced metabolism may explain the increased δ13C with latitudeobserved for adult edge portions (and marginally for juvenile coreportions). However, spatial variability was incorporated in this study,and even though a single cohort was analysed, spatial and temporalvariation in DIC or differences in diet may also explain the trends inδ13C with latitude.

Differences in δ13C between life stages were investigated withinindividuals to determine differences among latitude. If no movementthrough environments with different chemistries occurred with ontog-eny, dδ13C values would be similar among latitudes. Latitude, capturemonth and otolith weight influenced carbon isotopic signaturesthrough ontogeny, yet results provided little information to definepotential migration patterns. All individuals were enriched in δ13Cfrom juvenile to adult stages with less variation observed amongindividuals from 18 to 20°S, compared to individuals from 20 to 22°S.Greater variation in δ13C in the southmay be due to a number of causes,including the influence of temperature-inducedmetabolism (Høie et al.,2004), salinity (Elsdon and Gillanders, 2002), occurrence of differentoceanographic processes such as currents and upwelling events thataffect DIC (e.g. Patterson et al., 1999; Weeks et al., 2010), and/or reefgeography since the continental shelf is wider in the southern GBR.Simultaneous isotopic analyses of water are required to reveal thereasons for δ13C variation and elucidate links with movement patternsof L. miniatus.

To date, evidence for broad-scale movement of L. miniatus exists inthe few reports of long-distance movement (B. Sawynock; L. Currey;unpublished results) and hypotheses to explain regional differences inpopulation demographics (Williams et al., 2010). The current studyprovided evidence that different isotopic environments were inhabitedby themajority of individuals through ontogeny, via the analysis of δ18Oand δ13C in otoliths of adult individuals from different latitudes. Trendsin δ18O with latitude and life stage indicated potential for directional

migration of individuals, although further research is required to iden-tify how far individuals move between environments of differing isoto-pic signatures. This information demonstrates the potential for thistechnique to provide insight into movement at a scale important tothe management of exploited coral reef fishes.

Acknowledgements

The authors thank the James Cook University Cairns Analytical Unit,particularly Chris Wurster and Costjin Zwart for their guidance andexpertise with the laboratory isotope analysis. We acknowledge thoseinvolved in the previous ELF experiment and FRDC Project No. 98/131from which the samples were sourced. Map data was provided by theCommonwealth of Australia (Great Barrier Reef Marine Park Authority)and the authors thank Elodie Lédée for assistancewithmap production.Funding for this research was provided by the Australian ResearchCouncil (FT100101004) awarded to M.R.H. Additional research fundingwas granted to L.M.C. from James Cook University (JCU) School of Earthand Environmental Sciences (SEES). L.M.C. was also supported by anAIMS@JCU Postgraduate Research Scholarship and stipend. [ST]

References

Ayvazian, S.G., Bastow, T.P., Edmonds, J.S., How, J.R., Nowara, G.B., 2004. Stock structure ofAustralian herring (Arripis georgiana) in southwestern Australia. Fish. Res. 67 (1),39–53.

Bastow, T.P., Jackson, G., Edmonds, J.S., 2002. Elevated salinity and isotopic composition offish otolith carbonate: stock delineation of pink snapper, Pagrus auratus, in Shark BayWestern Australia. Mar. Biol. 141 (5), 801–806.

Bode, M., Bode, L., Armsworth, P.R., 2006. Larval dispersal reveals regional sources andsinks in the Great Barrier Reef. Mar. Ecol. Prog. Ser. 308, 17–25.

Breheny, P., Burchett,W., 2013. visreg: Visualization of RegressionModels. R Package Ver-sion 2.0-3.

Brown, I.W., Doherty, P., Ferreira, B., Keenan, C., McPherson, G.R., Samoilys, M., Sumpton,W.D., 1994. Growth, reproduction and recruitment of Great Barrier Reef food fishstocks. Report to Fisheries Research and Development Corporation, FRDC ProjectNo 90/18. Department of Primary Industries, Queensland.

Cadrin, C.X., Secor, D., 2009. Accounting for spatial population structure in stock assess-ment: past, present, and future. In: Beamish, R.J., Rothschild, B.J. (Eds.), The Futureof Fisheries Science in North America. Springer, pp. 405–426.

Campana, S.E., 1999. Chemistry and composition of fish otoliths: pathways, mechanismsand applications. Mar. Ecol. Prog. Ser. 188, 263–297.

Campana, S.E., 2005. Otolith elemental composition as a natural marker of fish stocks. In:Cadrin, S.X., Friedland, K.D., Waldman, J.R. (Eds.), Stock Identification Methods Appli-cations in Fishery Science. Academic Press, New York, pp. 227–245.

Carpenter, K.E., 2001. Lethrinidae. Emperor (emperor snappers). In: Carpenter, K.E., Niem,V. (Eds.), Bony Fishes Part 3 (Menidae to Pomacentridae), vol. 5. FAO, Rome, pp.3004–3051.

Cowen, R.K., 1990. Sex change and life history patterns of the labrid, Semicossyphuspulcher, across an environmental gradient. Copeia 1990 (3), 787–795.

Craig, H., Gordon, L.I., 1965. Deuterium and oxygen 18 variations in the ocean and marineatmosphere. In: Tongiogi, E. (Ed.), Stable Isotopes in Oceanographic Studies andPaleotemperatures. Spoleto, pp. 9–130.

Davies, C.R.,Williams, A.J., Mapstone, B.D., Benzie, J., van Herwerden, L., Choat, J.H., Adams,S., Murchie, C.D., Bean, K., Carlos, G., A.T., Ackerman, J., 2006. Stock structure and re-gional variation in population dynamics of the Red Throat Emperor and other targetspecies of the Queensland tropical reef line fishery. CRC Reef Research Centre Techni-cal Report No 61. CRC Reef Research Centre, Townsville.

Doherty, P.J., 2006. Variable replenishment and the dynamics of reef fish populations. In:Sale, P.F. (Ed.), Coral Reef Fishes: Dynamics and Diversity in a Complex Ecosystem.Academic Press, Burlington, pp. 327–355.

Dufour, V., Pierre, C., Rancher, J., 1998. Stable isotopes in fish otoliths discriminate be-tween lagoonal and oceanic residents of Taiaro Atoll (Tuaotu Archipelago, FrenchPolynesia). Coral Reefs 17, 23–28.

Edmonds, J.S., Fletcher, W.J., 1997. Stock discrimination of pilchards Sardinops sagax bystable isotope ratio analysis of otolith carbonate. Mar. Ecol. Prog. Ser. 152, 241–247.

Elsdon, T.S., Gillanders, B.M., 2002. Interactive effects of temperature and salinity on oto-lith chemistry: challenges for determining environmental histories of fish. Can. J. Fish.Aquat. Sci. 59 (11), 1796–1808.

Elsdon, T.S., Ayvazian, S., McMahon, K.W., Thorrold, S.R., 2010. Experimental evaluation ofstable isotope fractionation in fish muscle and otoliths. Mar. Ecol. Prog. Ser. 408,195–205.

Fairclough, D.V., Edmonds, J.S., Jackson, G.D., Lenanton, R.C.J., Kemp, J., Molony, B., Keay, I.S., Crisafuli, B.M., Wakefield, C.B., 2013. A comparison of the stock structures of twoexploited demersal teleosts, employing complementary methods of otolith elementanalysis. J. Exp. Mar. Biol. Ecol. 439, 181–195.

Fromentin, J.M., Powers, J.E., 2005. Atlantic bluefin tuna: population dynamics, ecology,fisheries and management. Fish Fish. 6 (4), 281–306.

25L.M. Currey et al. / Journal of Experimental Marine Biology and Ecology 456 (2014) 18–25

Gillanders, B.M., Able, K.W., Brown, J.A., Eggleston, D.B., Sheridan, P.F., 2003. Evidence ofconnectivity between juvenile and adult habitats for mobile marine fauna: an impor-tant component of nurseries. Mar. Ecol. Prog. Ser. 247, 281–295.

Grüss, A., Kaplan, D.M., Guénette, S., Roberts, C.M., Botsford, L.W., 2011. Consequences ofadult and juvenile movement for marine protected areas. Biol. Conserv. 144,692–702.

Hanson, N.N., Wurster, C.M., EIMF, Todd, C.M., 2010. Comparison of secondary ion massspectrometry and micromilling/continuous flow isotope ratio mass spectrometrytechniques used to acquire intra-otolith δ18O values of wild Atlantic salmon (Salmosalar). Rapid Commun. Mass Spectrom. 24, 2491–2498.

Høie, H., Folkvord, A., Otterlei, E., 2003. Effect of somatic and otolith growth rate on stableisotopic composition of early juvenile cod (Gadus morhua L.) otoliths. J. Exp. Mar. Biol.Ecol. 289 (1), 41–58.

Høie, H., Andersson, C., Folkvord, A., Karlsen, Ø., 2004. Precision and accuracy of stable iso-tope signals in otoliths of pen-reared cod (Gadus morhua) when sampled with ahigh-resolution micromill. Mar. Biol. 144, 1039–1049.

Huxam, M., Kimani, E., Newton, J., Augley, J., 2007. Stable isotope records from otoliths astracers of fish migration in a mangrove system. J. Fish Biol. 70, 1554–1567.

Jones, G.P., 1991. Post recruitment processes in the ecology of coral reef fish populations.In: Sale, P.F. (Ed.), The Ecology of Fishes on Coral Reefs. Academic Press, San Diego,pp. 294–328.

Kalish, J.M., 1991. Oxygen and carbon stable isotopes in the otoliths of wild andlaboratory-reared Australian salmon (Arripis trutta). Mar. Biol. 110, 37–47.

Kaunda-Arara, B., Rose, G.A., 2004a. Out-migration of tagged fishes from marine reef na-tional parks to fisheries in coastal Kenya. Environ. Biol. Fish. 70 (4), 363–372.

Kaunda-Arara, B., Rose, G.A., 2004b. Long-distance movements of coral reef fishes. CoralReefs 23, 410–412.

Leigh, G.M., Williams, A.J., Begg, G.A., Gribble, N.A., Whybird, O.J., 2006. Stock assessmentof the Queensland east coast red throat emperor (Lethrinus miniatus) fishery. Sustain-able Fisheries Unit, Animal ScienceDepartment of Primary Industries and Fisheries,Brisbane.

Mapstone, B.D., Davies, C.R., Little, L.R., Punt, A.E., Smith, A.D.M., Pantus, F., Lou, D.C.,Williams, A.J., Jones, A., Ayling, A.M., Russ, G.R., McDonald, A.D., 2004. The effects ofline fishing on the Great Barrier Reef and evaluations of alternative potential manage-ment strategies. CRC Reef Research Centre Technical Report No. 52. CRC Reef Re-search Centre, Townsville.

Moran, M., Burton, C., Jenke, J., 2003. Long-term movement patterns of continental shelfand inner gulf snapper (Pagrus auratus, Sparidae) from tagging in the Shark Bay re-gion of Western Australia. Mar. Freshw. Res. 54 (8), 913–922.

Nanami, A., Yamada, H., 2009. Site fidelity, size, and spatial arrangement of daytime homerange of thumbprint emperor Lethrinus harak (Lethrinidae). Fish. Sci. 75 (5),1109–1116.

Newman, S.J., Buckworth, R.C., Mackie, M.C., Lewis, A.R., Wright, I.A., Williamson, P.C.,Bastow, T.P., Ovenden, J.R., 2009. Spatial subdivision among assemblages of Spanishmackerel, Scomberomorus commerson (Pisces: Scombridae) across northernAustralia: implications for fisheries management. Glob. Ecol. Biogeogr. 18, 711–723.

Newman, S.J., Wright, I.A., Rome, B.M., Mackie, M.C., Lewis, P.D., Buckworth, R.C., Ballagh,A.C., Garrett, R.N., Stapley, J., Broderick, D., Ovenden, J.R., Welch, D.J., 2010. Stockstructure of Grey Mackerel, Scomberomorus semifasciatus (Pisces: Scombridae) acrossnorthern Australia, based on otolith stable isotope chemistry. Environ. Biol. Fish. 89,357–367.

Newman, S.J., Pember, M.B., Rome, B.M., Mitsopoulos, G.E.A., Skepper, C.L., 2011. Stockstructure of blue threadfin Eleutheronema tetradactylum across northern Australiaas inferred from stable isotopes in sagittal otolith carbonate. Fish. Manag. Ecol. 18,246–257.

Patterson, H.M., Thorrold, S.R., Shenker, J.M., 1999. Analysis of otolith chemistry in Nassaugrouper (Epinephelus striatus) from the Bahamas and Belize using solution-based ICP-MS. Coral Reefs 18 (2), 171–178.

Pinheiro, J., Bates, D., DebRoy, S., Sarkar, D., R Core Development Team, 2013. nlme: Linearand Nonlinear Mixed Effects Models. R Package Version 3.1-109.

Pittman, S.J., McAlpine, C.A., 2001. Movements of marine fish and decapod crustaceans:process, theory and application. Adv. Mar. Biol. 44, 205–294.

Sale, P.F., 1991. The Ecology of Fishes on Coral Reefs. Academic Press, New York.Schwarcz, H.P., Gao, Y., Campana, S., Browne, D., Knyf, M., Brand, U., 1998. Stable carbon

isotope variations in otoliths of Atlantic cod (Gadus morhua). Can. J. Fish. Aquat. Sci.55, 1798–1806.

Steer, M.A., Halverson, G.P., Fowler, A.J., Gillanders, B.M., 2010. Stock discrimination ofSouthern Garfish (Hyporhamphus melanochir) by stable isotope ratio analysis of oto-lith aragonite. Environ. Biol. Fish. 89, 369–381.

Taylor, B.M., Mills, J.S., 2013. Movement and spawning migration patterns suggest smallmarine reserves can offer adequate protection for exploited emperorfishes. CoralReefs. http://dx.doi.org/10.1007/s00338-013-1065-6.

Thorrold, S.R., Campana, S.E., Jones, C.M., Swart, P.K., 1997. Factors determining δ13C andδ18O fractionation in aragonitic otoliths of marine fish. Geochim. Cosmochim. Acta 61(14), 2909–2919.

Thorrold, S.R., Jones, C.M., Swart, P.K., Targett, T.E., 1998. Accurate classification of juvenileweakfish Cynoscion regalis to estuarine nursery areas based on chemical signatures inotoliths. Mar. Ecol. Prog. Ser. 173, 253–265.

Tobin, A., Schlaff, A., Tobin, R., Penny, A.L., Ayling, T., Ayling, A., Krause, B., Welch, D.J.,Sutton, S., Sawynok, B., Marshall, N., Marshall, P., Maynard, J., 2010. Adapting tochange: minimising uncertainty about the effects of rapidly-changing environmentalconditions on the Queensland Coral Reef Fin Fish Fishery. Final Report to the FisheriesResearch & Development Corporation, Project 2008/103. Fishing & Fisheries ResearchCentre Technical Report No. 11. James Cook University, Townsville.

Weeks, S., Bakun, A., Steinberg, C., Brinkman, R., Hoegh-Guldberg, O., 2010. The CapricornEddy: a prominent driver of the ecology and future of the southern Great Barrier Reef.Coral Reefs 29 (4), 975–985.

Weidman, C.R., Millner, R., 2000. High-resolution stable isotope records from NorthAtlantic cod. Fish. Res. 46 (1–3), 327–342.

Williams, D.M., Russ, G.R., 1994. Review of data on fishes of commercial and recreationalfishing interest on the Great Barrier Reef. Report to the Great Barrier Reef Marine ParkAuthority. Great Barrier Reef Marine Park Authority, Townsville, p. 103.

Williams, A.J., Davies, C.R., Mapstone, B.D., Russ, G.R., 2003. Scales of spatial variation indemography of a large coral-reef fish — an exception to the typical model? Fish.Bull. 101 (3), 673–683.

Williams, A.J., Davies, C.R., Mapstone, B.D., 2005. Variation in the periodicity and timing ofincrement formation in red throat emperor (Lethrinus miniatus) otoliths. Mar.Freshw. Res. 56 (5), 529–538.

Williams, A.J., Davies, C.R., Mapstone, B.D., 2006. Regional patterns in reproductive biologyof Lethrinus miniatus on the Great Barrier Reef. Mar. Freshw. Res. 57 (4), 403.

Williams, A.J., Mapstone, B.D., Davies, C.R., 2007a. Spatial and interannual patterns ingrowth of an exploited coral-reef fish. J. Fish Biol. 71 (4), 970–992.

Williams, A.J., Mapstone, B.D., Davies, C.R., 2007b. Spatial patterns in cohort-specific mor-tality of red throat emperor, Lethrinus miniatus, on the Great Barrier Reef. Fish. Res.84, 328–337.

Williams, A.J., Little, L.R., Punt, A.E., Mapstone, B.D., Davies, C.R., Heupel, M.R., 2010. Ex-ploring movement patterns of an exploited coral reef fish when tagging data are lim-ited. Mar. Ecol. Prog. Ser. 405, 87–99.

Wurster, C.M., Patterson, W.P., Cheatham, M.M., 1999. Advances in micromilling tech-niques: a new apparatus for acquiring high-resolution oxygen and carbon stable iso-tope values and major/minor elemental ratios from accretionary carbonate. Comput.Geosci. 25, 1159–1166.

Zuur, A.F., Ieno, E.N., Walker, N., Saveliev, A.A., Smith, G.M., 2009. Mixed Effects Modelsand Extentions in Ecology with R. Springer.