Oscillatory zinc distribution in Arctic char (Salvelinus alpinus) otoliths

ELSEVIER Ftsherles Research 46 (2000) 289-298

www.elsevier.com/locate/fishres

Oscillatory zinc distribution in Arctic char(Salvelinus alpinus) otoliths:

The result of biology or environment?

Norman M. Halden”‘*, Sergio R. Mejia”, John A. Babalukb, James D. Reistb,Allan H. Kristoffersonb, John L. Campbell”, William J. Teesdale”

“Department of Geological Sciences. University of Munitoba, Winnipeg, Man., Cunada R3T 2N2‘Canada Department of Fisheries and Oceans, 501 University Crescent. Winnipeg, Man., Canado R3T 2N6

‘Department of Physics, University of Guelph, Guelph, Ont., Canada NIG 2WI

Abstract

Scanning proton microprobe (SPM) analysis and imaging was used to map the distribution of zinc in otoliths of

anadromous and non-anadromous Arctic char (Salvelinus alpinus) from the Canadian Arctic. Zinc distribution patterns were

oscillatory with concentrations ranging from 35 to 240 ppm. Superimposition of the zinc distribution on optical images of theotoliths permitted correlation of zinc uptake with annular structure in the otoliths and with strontium patterns from the sameotolith. Well defined oscillations of zinc concentrations were observed in the otoliths, particularly in the first few years in allsamples. In anadromous fish, these overlapped with the onset of the strontium oscillations that were indicative of anadromyand then declined; in non-anadromous fish the oscillations generally continued to later years. Oscillatory zoning of zinc mayindicate: (1) variations in the concentration of zinc in water; (2) variations in nutrient availability in the environment; (3)temperature variations in the habitat occupied; or (4) a combination of zinc concentration, nutrient availability or temperaturechanges. As such, the systematic distribution of zinc in otoliths has the potential to provide temporally constrained

information on fish habitat and/or fish biology. c 2000 Elsevier Science B.V. All rights reserved.

Keywords Zinc, Micro-PIXE; Trace-element; Anadromy; Nutrient supply

1. Introduction

Otoliths, the calcified structures found in the inner

ear of teleost fish, are constructed of alternating layers

of aragonite and protein which are thought to be

deposited annually (Degens et al., 1969). The annular

*Corresponding author. Tel.: + I-204-474-8857; fax: + I-204.

474-1623.

E-rwil c~&fwss: nn-halden @umanitoba.ca (N.M. Halden)

structure of the otolith thus provides a temporal record

of the growth history of the fish. Otoliths are routinely

used to age fish (e.g., Chilton and Beamish, 1982).

More recently chemical analysis of otoliths, besides

being used to assess aspects of fish behaviour, has also

been used to characterise the environment which the

fish occupied (Kalish, 1989; Radtke, 1989; Halden

et al., 1996; Babaluk et al., 1997). From a biological

perspective, age information and chemical records are

important because they provide valuable information

Ol65-7836/00/$ - see front matter i(‘, 2000 Elsevier Science B.V. All rights reserved

PII: SO 165.7X36(00)00154-5

2 9 0 NM. Halden et al./Fisheries Research 46 (2000) 289-298

that is necessary to develop conservation strategies fordifferent fish species. However, in addition to this,chemical information from otoliths has the potential tobe used to assess the productivity or the health of anaquatic environment annually throughout the life ofthe fish (i.e., represents a historical record).

Otoliths have the potential to contain many ele-ments that will substitute for Ca*+ in the aragonitecrystal structure (cf. Campana et al., 1997; Haldenet al., 1998). A number of studies have shown thatotoliths can absorb strontium and barium as well astransition elements (principally iron, manganese andzinc; e.g., Sie and Thresher, 1992; Halden et al., 1998).The key here is to link the elemental absorption to theannular structure of the otolith to provide temporalinformation on the fish’s biology and/or environment.From a mineralogical perspective, the material ofinterest is calcium carbonate (CaC03) in the structuralform of aragonite and here much is already knownabout the crystal chemistry of aragonite as well asmethods appropriate for its analysis (Reeder, 1983;Veizer, 1983). Although vaterite and calcite have beenreported, neither of these two structural polymorphswere observed in the otoliths in this study. A secondissue deriving ultimately from mineralogy, is that thechemistry of the fish’s environment is likely to beinfluenced by the mineralogy, weathering, drainageand geochemical character of the rocks surroundingthe waters in which the fish lives.

Annual growth increments of aragonite in otolithsare typically about 50-100 pm wide. In reflected lighteach annulus consists of a wide light region corre-sponding to a period of rapid summer growth (whichhas been tentatively linked with a more nutrient-richand warmer environment) and a thin dark region that isacquired during the winter (i.e., representing slowergrowth in a colder environment). Occasionally, thinnergrowth increments can be observed which are thoughtto correspond to daily additions of aragonite. How-ever, these zones are on the order of 2-5 u wide andare too small to be distinguished and analysed usingcurrent microbeam technology.

Chemical analysis of otoliths using proton-inducedX-ray emission (PIXE) has to date been primarilyaimed at determining strontium content. Thisapproach is non-destructive and has the necessaryspatial resolution and sensitivity to analyse annulifor elements such as strontium, barium, iron, manga-

nese and zinc in the carbonate matrix to the low partsper million level (ca. 2-5 ppm; Campbell et al., 1995).Micro-PIXE has been successfully used to distinguishbetween migratory and non-migratory char revealingmarked changes in strontium content correspondingto migrations between freshwater and marine envir-onments (Halden et al., 1995, 1996; Babaluk et al.,1997). Otoliths from non-migratory Arctic charfrom the Canadian Arctic have characteristically flatstrontium profiles throughout life, whereas otolithsfrom migratory char have flat and comparativelylow-strontium levels in the core area and the firstseveral annuli (indicating a period of freshwater resi-dence) followed by a sudden increase in the strontiumcontent as well as the appearance of a distinctiveoscillatory zonation in strontium concentrations whichcorresponds to migrations between freshwater andmarine environments. These strontium profiles areused to reference the onset of migratory behaviourin this study.

While in the past the main focus of microbeam workhas been the analysis of strontium, Coote et al. (1991)and Sie and Thresher (1992) reported zinc in someotoliths, but they were unable to determine the detailsof the zinc distribution. The ultimate source of zinc inthe environment has to be the lithosphere, particularlythe rocks surrounding the water in which the fish lives.However, the availability and uptake of zinc by fishappears to be dependent upon the specific environmentoccupied, as well as on physiological processes occur-ring in the fish. Willis and Sunda (1984) argue that insea water the primary mode of zinc uptake appears tobe via food. No similar assessment appears to havebeen done in fresh water. In addition, zinc may well beabsorbed by the fish through active ion transportthrough the gills. The availability of zinc, food-chainbiomagnilication, and the flux of zinc depends uponthermal regime, pH, alkalinity and amount of parti-culate matter in the water, all of which vary with thetype of aquatic environment. If zinc shows a systema-tic distribution in otoliths related to annular structure,analysis of this element may provide, in addition tostrontium, temporally constrained information onhabitat, fish behaviour or nutrient supply. This studypresents the results of scanning proton microprobe@PM) analysis of zinc in Arctic char otoliths andcompares zinc distribution to the distribution of stron-tium and the optical structure of the otolith.

N.M. Halden et al. /Fisheries Research 46 (2ooO) 289-298 291

2. Materials and methods 2.2. Sample preparation and analytical methods

2.1. Arctic char samples from northern Canada

Otoliths were taken from (1) known anadromouschar from Halovik River (the fish were sampledreturning from the sea); (2) spawning anadromouschar from Halovik Lake; (3) two forms of non-ana-dromous char from Lake Hazen; and (4) char fromCraig Lake (Pig. 1). The population of char from LakeHazen includes two morphotypes, large and small char(Reist et al., 1995). A previous study (Johnson, 1983)assumed that large-form fish were anadromous, how-ever, examination of otolith strontium demonstratedthat both morphotypes remain; in freshwater for theirentire life (Babaluk et al., 1997). This has been sub-sequently supported by a radio-telemetry study (J.Babaluk, unpublished data). The population in CraigLake shows morphological characteristics of small,land-locked (non-migratory) char despite an indirectconnection to the sea. Previous studies of otolithstrontium have also confirmed the non-anadromy ofthis population (Babaluk et al., 1997).

The methodology used to mount the otolith issimilar to that used in the preparation of small mineralsamples for electron microprobe analysis (EMPA). Alongitudinal section of the otolith is embedded in anepoxy resin within a leucite-disc probe mount(25.4 mm diameter). The surface of the disc andotolith are ground such that the core is exposed andthe ground surface is then polished. As aragonite isessentially a light matrix with regard to proton pene-tration and X-ray absorption, sufficient thickness ofmineral (>60 um) must be retained to stop the protonbeam in order that excitation of the mounting mediumis avoided.

A high-resolution monochrome image of the otolithis collected using a Kontron image analysis systemconnected to a reflection microscope using glancingreflected light from a fibre optic light source; thisprovides for clear optical contrast and differentiationof the annuli. imaging is done before carbon coating ascarbon coating tends to degrade the optical images. Oneach image, the starting and end points for the PIXE

Lake Hazen 7 5

Qmenland

Fig. 1. Locations of Canadian Arctic char populations from which the otoliths were collected for scanning proton microprobe analysis.

292 NM. Halden et al. /Fisheries Research 46 (2000) 289-298

line-scan are defined within the first annulus (corearea) and the otolith edge, respectively. The line islocated such that, as far as is possible, the line-scan isperpendicular to the annuli. The final step, in prepara-tion for probe work, requires that the otolith is carboncoated to prevent charging under the beam.

One-dimensional line-scans are elemental mapsobtained by appropriately rastering the proton beam(5 x5 pm at 3 MeV); X-ray intensities and corre-sponding X-Y coordinates are recorded on computerdisc in list mode. The sum spectrum of all recorded X-ray events is assembled, and energy windows for thetrace elements of interest are defined. The sum spec-trum is fitted by the GUPIX software (Maxwell et al.,1989) thereby providing the continuum backgroundintensity in each window per pC of proton charge.

On completion of the line-scan, the otolitb is re-imaged to verify the location of the proton beamrelative to the annuli; this can be seen as a discoloura-

tion of the carbon coat (residual hydrocarbon stain)and as a line in the epoxy where the beam crosses theedge of the otolith. The PIXE line-scan file for theelement(s) of interest is imported into the imageanalyser. The start of the file is identified with theselected scan origin in the core area, and the end isidentified in the line-scan by the disappearance of thecalcium signal as the beam moves from the otolithedge to epoxy. The optical image and the concentra-tion scans are then scaled to have the same lateralextent in a compound image.

For point analyses, a 5 x 5 pm 3 MeV proton beamwas used with average beam-currents of 6-l nA andthe time integrated charge for each analysis was2.5 uC; X-ray spectra were recorded with a Si (Li)detector. The X-rays of the trace elements of interestoccur in the energy region 6-14 keV, so a combinationof 0.125 urn mylar and 106 urn aluminium filters wasused to suppress the intensity of the lower-energy

Fig. 2. Combined otohth and line-scan image from an anadromous Halovik River Arctic char (fork length=660 mm, weight=3201 g, male,

age= 13+ years). The zinc line-scan has been superimposed on the proton beam path (seen as a residual hydrocarbon stain on the inset image

of the carbon coated otolith). For comparison, the strontium line-scan from the same path is shown above the otolith. Annuli along the line-

scan are indicated by the triangles, years 1, 5, 10 and 13 are labeled for reference. All following figures are presented in a similar manner.

NM Halden et al. /Fisheries Research 46 (2ooO) 289-298 293

region of the spectra; this reduces the calcium X-rays line-scans for zinc and strontium superimposed; thesereaching the detector and gives better counting sta- are shown as X-ray intensity for each element. Thetistics for the trace elements. Spectra were processed zinc peaks are the same width as the annuli, moreover,using GUPIX (Maxwell et al., 1989) which extracts there is a spatial correspondence of the zinc peaks withpeak areas from a spectrum using a non-linear least- the annuli; the zinc peaks correspond to the lighter-squares-fitting procedure. A synthetic spectrum is coloured part of the annuli. This spatial correspon-derived using estimated concentrations of elements dence is also preserved where zinc peaks and stron-of interest and a database which includes relative X- tium peaks coincide with the annuli in the outer part ofray intensities. The background is removed via a top- the otolith where migratory behaviour is recorded. Thehat filter method. The calculations for the element strontium scan is typical of what is seen in otolithsconcentrations are then iterated until the best fit taken from anadromous char. It has low and relativelybetween the synthetic and observed spectrum is constant strontium concentrations for the core areaattained. Further details on spectrum fitting and cali- and first several annuli followed in the outer annuli by

bration procedures are given by Campbell et al. (1995) a sudden increase in strontium concentrations and the

and Campana et al. (1997). development of a distinctly oscillatory distribution.

3. Results

Fig. 2 shows a typical grey-level reflected lightimage of a Halovik River Arctic char otolith with

Fig. 3 shows line-scans on an otolith collected froma fish sampled from Halovik Lake. The strontium scanshows the char was anadromous. The image showssome strontium variation in the first four annulibeyond the core but this is at a much lower concen-tration of strontium compared to what is seen in the

Fig. 3. Combmed otohth and line-scan image from an anadromous Halovik Lake Arctic char (fork length=803 mm. weight=5 191 g, male,

age= I2+ years)

294 NM Halden et al./Fisheries Research 46 (2000) 289-298

outer annuli. The zinc distribution shown in Fig. 3 isalso oscillatory and the zinc peaks are again similar inwidth to the annuli with the peaks coinciding with thelight part of the annulus. The overall concentration ofzinc is higher in the core and the first several annuli.Annuli l-8 have elevated oscillatory zinc concentra-tions; annuli 7 and 8 also show elevated strontiumconcentrations indicating that there is some overlap ofthe zinc signature with the onset of anadromy.

It is worth noting here that Fig. 2 shows an otolithtaken from a fish caught in Halovik River at tbe sametime as the Halovik Lake catch. The structure of, andrelationship between, the zinc and strontium patternsare similar with all the prominent zinc peaks (asso-ciated with broad distinct annuli) occurring in the first5-6 years. This is followed by a more extended periodwhere there is overlap in the oscillatory zinc andstrontium signals. In the outermost part of the otolith,zinc declines but the strontium retains its prominentoscillatory pattern.

Fig. 4 shows a grey-level reflected light image of anotolith collected from a fish from Craig Lake onEllesmere Island. The fish’s life history is interpretedto be non-anadromous (Craig Lake is not directlyconnected to the sea, Babaluk et al., 1997); the imagealso includes the superimposed SPM scans for zincand strontium. This image is qualitatively differentfrom those seen in Figs. 2 and 3. The strontium scan isessentially flat, consistent with no excursions to amarine environment, on the other hand, the zincdistribution is oscillatory throughout the fish’s life.Again the zinc peaks are of similar width to the annuliand there is a coincidence of the peak with the lightpart of the annulus.

Line-scans were also taken from otoliths of bothsmall- and large-form Arctic char (23 in total) fromLake Hazen, Ellesmere Island (Figs. 5 and 6, respec-tively). In all cases the strontium patterns show novariation indicating that neither of the morphotypes goto sea. However, in both cases the zinc patterns show

Rg. 4 Combmed otoltth and line-scan image from a non-anadromous Craig Lake Arctic char (fork length=240 mm, weight= 121 g. male,

age= IO+ years).

.

N.M. Halden et al./Fisheries Research 46 (2ooO) 289-298 295

Fig. 5. Combined otolitb and line-scan image from a non-anadromous small-form Lake Hazen Arctic char (fork length=360 mm,weight=476 g, male, age=25+ years). The line-scan was reoriented to ensure the proton beam crossed all annuli. The image has been

reconstructed to reflect this. A similar presentation is adopted for Fig. 6.

Fig. 6. Combined otolith and line-scan image from a non-anadromous large-form Lake Hazen Arctic char (fork length=569 mm,weight=2037 g, male, age=20+ years).

296 NM. Halden et al./Fisheries Research 46 (2000) 289-298

Table 1Summary of otolith point analytical data showing zinc content ofthe core area

Population

Halovik RiverHalovik LakeLake Hazen, large-formLake Hazen, small-formCraig Lake

No. ofotoliths/No.of analyses

S/30513013fl8lo/604l24

Mean zincconcentration

(ppm)* 1

57f1993f3385f3467f2369f29

distinct oscillations corresponding to the early years ofthe fish’s life, followed by a declining zinc signal. Inthe large morphotype, the oscilIatory nature of the zincsignal remains well defined throughout the fish’s lifebut this is not the case for the smaller morphotype.While this latter observation appears to be in contrastto Craig Lake char, it has to be noted that the typicalage of the Lake Hazen fish sampled was ca. 20 years,whereas those from Craig Lake were typically ca. 10years old.

Point analyses were done in the core areas of theotoliths to assess the absolute concentrations of stron-tium and zinc and these are summarised for Zn inTable 1; this indicates that zinc is easily detected in theotoliths above the 2 ppm detection limit (cf. Campbellet al., 1995). Zinc ranges from about 30-140 ppm inthe otoliths. Strontium has the potential to be used todiscriminate the source lakes for fish (Halden et al.,1996), on the other hand, there is considerable overlapin the absolute zinc concentrations (except for HalovikRiver) thus, zinc is unlikely to be particularly usefulfor discriminating groups of fish. However, the varia-tion is quite different for each locality. This mayreflect the amplitude of the zinc oscillations, variationsin the zinc availability in the local environment, orgrowth rate.

4. Discussion

Optical microscopy is routinely used to determinethe ages of fish by counting annuli on calcified struc-tures and this approach can be supplemented in somecases by high-resolution probing of otolith chemistryto discern life-history events. For example, analysis ofstrontium distribution has been used to distinguish

migratory from non-migratory Arctic char, which hasled to a better understanding of the behaviour and life-histories of char from Lake Hazen (Babaluk et al.,1997).

Arctic char otoliths contain distinctive oscillatoryzinc patterns with some consistent features: (1) thezinc distribution is oscillatory and corresponds to theannular structure of the otolith which means that thefish’s uptake of zinc follows a yearly pattern; (2) themost prominent zinc peaks seem to occur in the firstseveral annuli indicating that the most rapid uptake ofzinc is occurring in the early years of the fish’s life; (3)the oscillatory zinc pattern is seen in both anadromousand non-anadromous fish; (4) in general, the oscilla-tory zinc declines in the outer annuli suggesting thatthe uptake of zinc is decreasing and something ineither the fish’s behaviour, metabolism, and/or envir-onment has changed and this is consistent regardlessof the specific environment or life history of the fishand (5) in anadromous fish, there is some overlapbetween the end of the oscillatory zinc signal and theonset of the oscillatory strontium signal; however, themost prominent zinc peaks seem to occur before theonset of anadromy.

In the case of the anadromous fish, an idealisedinterpretation of these patterns might be (1) anapproximately O-6 or 7 year freshwater phase wherestrontium is low and zinc is available for cyclicaluptake dependent possibly on food supply, tempera-ture, and alkalinity; (2) an approximately 2-3 yearphase encompassing the smoltification phase wherethe fish periodically encounters a brackish (perhapsestuarine environment) where strontium is increasingand zinc is declining; and (3) a migratory adult phasecorresponding to where the bulk of feeding is beingdone in a high-strontium, low-zinc marine environ-ment followed by a return to freshwater; here, how-ever, with no significant feeding there will be littleuptake of zinc and strontium is characteristically low.

There also appears to be a relationship between thezinc content and size in the case of the fish from LakeHazen. There is a longer and more distinct oscillatoryzinc signal in the large-form char otoliths. Given thatlarge and small forms exist in the same environment itis possible that the greater zinc content could berelated to diet or metabolism.

Studies aimed at understanding the concentration oftrace elements in marine organisms show that food is a

NMNM.n. Halden et al. /Fisheries Research 46 (2ooO) 289-298 297

significant factor in the overall uptake of elementssuch as zinc from the environment (e.g., Renfro et al.,1975; Cossa et al., 1980; Milner, 1982; Willis andSunda, 1984). Other studies have shown the effects ofzinc absorption through gills and related this to toxi-city, water pH and alkalinity (e.g., Bradley and Spra-gue, 1985; Campbell and Stokes, 1985; I&k et al.,1995 ). An additional feature of some of these studiessuggests that there is a correlation between weight(and presumably the size) of the organism and itstrace-element content. For example, in a study of O-group plaice, Milner (1982) showed that where thezinc uptake was from water, those fish with the highestzinc concentration had the lowest metal content andsize accounted for the concentration differences.Bradley and Sprague (1985) observed that increasedsize correlated with reduced zinc tolerance. Bothstudies went on to suggest that metabolic regulationgoverned the uptake of zinc. It is worth noting that allthe patterns in this study show an overall decrease inthe zinc content with age (which may in fact be relatedto size) and that maybe there is a declining metabolicneed for zinc as the fish get older and larger.

Many of these studies while demonstrating a likelyzinc pathway into the fish were aimed at understand-ing toxic trace-metal (including cadmium and lead)uptake in contaminated environments. In general, it isimportant to distinguish natural zinc signals from whatmight be considered contamination. In the areas fromwhich our fish were sampled there are no anthropo-genic sources of zinc. If there were an anthropogenicinfluence, then in order for it to have influenced such awide region over which the fish were taken, the zincwould have to have been airborne and coherently andannually variable over a period of 25 years (the oldestotoliths in the study), which is very unlikely.

It is much more likely that the zinc content of theotolith reflects the zinc availability to the fish eitherwithin its local environment or through metabolism. Ifzinc was being obtained primarily through food then itmay be possible to interpret the oscillatory signal as, inpart, a seasonal signal. Here, summers with presum-ably elevated temperatures would correspond to per-iods of rapid nutrient production and uptake producingprominent zinc peaks, whereas winters would corre-spond to periods of little nutrient uptake with the zincdeclining to background levels. It is worth notingagain that the zinc peaks correspond to the wider

lighter-coloured parts of the annuli which reflect themore rapid (summer) growth period of the otolith. Bylinking the zinc uptake to diet, the overall pattern mayreflect size, in a general semi-quantitative way.

If the zinc signal can indeed be linked to the size ofthe fish (through their nutrient intake) and the pro-ductivity of the environment, then the patterns maycontain interesting information on the timing of life-history events. In the case of the Halovik Lake charthey all appear to have acquired significant zinc in thefirst 5 or 6 years of their lives. If Halovik Lake is aproductive environment, the fish may have reached anadequate size to migrate by the fifth year of their lives.This would be intermediate given the range of firstmigration to sea (2-9 years) identified by Johnson(1980). In this way the productivity of an environmentmay influence when fish migrate; moreover, if theproductivity is linked to local weather, the annularrecord of the otolith may also be linked in a generalway to climate.

SPM and image analysis makes it is possible torelate the distribution of trace elements to the annularstructure of otoliths. With many elements, it may bepossible to more specifically characterise fish beha-viour, environment and population stock structure.SPM is an ideal tool for this purpose and its use, incombination with other techniques (e.g., laser ablationmass spectrometry) should make it possible to addresssuch characterisation in detail.

Acknowledgements

The project was supported by the Canada Depart-ment of Fisheries and Oceans. JLC and NMH aresupported by the Natural Sciences and EngineeringResearch Council of Canada. We deeply appreciatethe interest and the financial and logistical supportprovided by the Department of Canadian Heritage(Parks Canada), the Department of Natural ResourcesCanada (Polar Continental Shelf Project), and theTungavik Federation of Nunavut.

References

Babaluk, J.A., Halden, N.M., Reist. J.D., Kristofferson, A.H.,Campbell, J.L., Teesdale, W.J., 1997. Evidence for non-anadromous behaviour of Arctic char (Salvelinus alpinus) from

298 N.M. Ha&n el al./Fisheries Research 46 (2&W) 289-298

Lake Hazen, Bllesmere Island, Northwest Territories, Canada,based on otolith strontium distribution. Arctic 50, 223-224.

Bradley, R.W., Sprague, J.B., 1985. Accumulation of zinc byrainbow trout as influenced by pH, wabr hardness and fish size.Envion. Toxicol. 4, 685-694.

Campana. S.E., Tborrold, S.R., Jones, C.. Gunther, D., Turbrett,M.. Longerich, H.. Jackson, S., Halden, N.M., Kalish. J.M..Piccoli. P. de Pontual, H., Troadec, H., Panfilli, J.. Secor, D.H.,Severin, K., Sie, S.H.. Thresher, R., Campbell, J.L., Teesdale.W.J., 1997. Comparison of accuracy, precision and sensitivityin elemental analysis of fish otoliths using the electronmicroprobe., PIXE and laser ablation ICPMS. Can. J. Fish.Aquat. Sci. 54, 2068-2079.

Campbell, J.L., Teesdale, W.J., Halden, NM., 1995. Theory,practice and application of PIXE microanalysis and SPMelement mapping. Can. Mineral. 33, 279-292.

Campbell, P.G.C., Stokes, P.M., 1985. Acidification and toxicity ofmetals to aquatic biota. Can. J. ysh. Aquat. Sci. 42, 2034-2049.

Chilton, D.E., Beamish, R.J., 1982. Age determination methodsfor fishes studied by the Groundfish Program at the PacificBiological Station. Can. Spec. Pub]. Fish. Aquat. Sci. 60,102 pp.

Coute, G.E., Gauldie, R.W., West, I.E. 1991. Application of anuclear microprobe to the study of fish otolitbs and scales.Nucl. Instr. Meth. B 54, 144-150.

Cossa D., Bomger, E.. Pouliot. D., Piuze, J.. Chanut, JP., 1980.Geographical and seasonal variations in the relationshipbetween trace metal content and body weight in Mytillcr edulis.Mar. Biol. 58, 7-14.

Degens, E.T., Deuser, W.C., Haedrich, R.L., 1969. Molecularstructure and composition of fish otoliths. Mar. Biol. 2, 105-113.

Halden, N.M., Babaluk, J.A., Campbell, J.L., Teesdale, W.J., 1995.Scanning proton microprobe analysis of strontium in an Arcticchar, Salvelinus a&inus, otolith: implications for the inter-pretation of anadromy. Envrion. Biol. Fish. 43, 333-339.

Halden. N.M., Babaluk, J.A., Kristofferson, A.H., Campbell, J.L..Teesdale, W.J., Maxwell, J.A., Reist. J.D., 19%. Micro-PIXEstudies of Sr zoning in Arctic char otoliths: migratorybehaviour and stock discrimination. Nucl. Instr. Meth. B 109/110, 592-597.

Halden. N.M., Mathers, K., Fryer, B.J., Chen, X., 1998. Preserva-tion of historical records of aqueous geochemistry in biogenic

carbonates. GAUMAC Annual Meeting, Quebec City, May18-20, program with abstracts, pp. A70.

Johnson, L., 1980. The Arctic charr, Salvelinus alpinus. In: Balon,E.K. (Ed.). Charm: Salmonid Fishes of the Genus &lvelinu.r.Dr. W. Junk Publishers, The Hague, pp. 15-98.

Johnson. L., 1983. Homeostatic characteristics of single speciesfish stocks in Arctic lakes. Can. J. Fish. Aquat. Sci. 40, 987-1024.

Kalish, J.M., 1989. Otolith microchemistry: Validation of theeffects of physiology, age and environment on otolithcomposition. J. Exp. Mar. Biol. Ecol. 132, 151-178.

K&k. G., Hofer, R., Wdgrath, S., 1995. Accumulation of tracemetals (Cd, F’b, Cu. Sr) in Arctic char (Salvelinus alpine) fromoligotrophic Alpine lakes: relation to alkalinity. Can. J. Fish.Aquat. Sci. 52, 2367-2376.

Maxwell, J., Campbell, J.L., Teesdale, W.J., 1989. The GuelphPIXE software package. Null. Instr. Metb. B 43, 218-230.

Milner, N.J., 1982. The accumulation of Zn by O-group plaice,Pleumnectes plantessa (L.), from high concentration in seawater and food. J. Fish Biol. 21, 325-336.

Radtke, R.L., 1989. Strontium-calcium concentration ratios in fishotoliths as environmental indicators. Comp. B&hem. Physiol.92 (A). 189-193.

Reeder, R.J., 1983. Crystal chemistry of the rhombohedralcarbonates. In: Reeder, R.J. (Ed.), Carbonates: Mineralogyand Chemistry. Rev. Mineral. 11, 147.

Reist, J.D., Gyselman. E., Babaluk, J.A., Johnson, J.D.. Wissink,R., 1995. Evidence for two morphotypes of Arctic char(Salvelinus alpinw (L.)) from Lake Hazen, Ellesmere Island,Northwest Territories, Canada. Nordic J. Freshwater Re.s. 71,396-410.

Renfro, W.C., Fowler, S.W., Heyraud, M., La Rosa, J., 1975.Relative importance of food and water in long-term Zi#accumulation in marine biota. J. Fish. Res. Board Can. 32,1339-1345.

Sie, S.H., Thresher, R.E., 1992. Micro-PIXE analysis of fishotolitbs: methodology and evaluation of f& results for stockdiscrimination. Int. J. PIXE 2, 357-379.

Veizer, J., 1983. Trace elements and isotopes in sedimentarycarbonates. In: Re=eder, R.J. (Ed.), Carbonates: Mineralogy andChemistry. Rev. Mineral. 11. 265-300.

Willis, J.N., Sunda, W.C., 1984. Relative contributions of food andwater in the accumulation of zinc by two species of marine fish.Mar. Biol. 80, 273-279.