Fitzgerald et al.: Elemental signatures in otoliths of larval Theragra chakogramma 



605 



ineffective because of the low level of exchange required 

 to maintain genetic homogeneity, at least over ecological 

 time scales, and the low level of genetic drift associated 

 with large populations (Waples, 1998; Hellberg et al., 

 2002). However, preliminary studies have indicated that 

 otolith geochemistry may prove to be a useful natural 

 tag of population structure in walleye pollock (Severin 

 et al., 1995). Otoliths are accretionary crystalline struc- 

 tures located within the inner ear of teleost fish. They 

 are formed through concentric additions of alternating 

 protein and aragonite layers around a central nucleus. 

 The use of otoliths as natural geochemical tags is con- 

 tingent on the metabolically inert nature of the otolith 

 and the fact that once deposited, otolith material is 

 neither resorbed nor metabolically reworked (Campa- 

 na and Neilson, 1985; Campana, 1999). The chemical 

 composition of otoliths also reflects to some degree the 

 physicochemical characteristics of the ambient water 

 (Bath et al., 2000). If the water where pollock reside 

 has distinct oceanographic characteristics, then many of 

 the elements incorporated into the otoliths should differ 

 among locations. Migrations between water masses at 

 some age will, therefore, be recorded in the chemical 

 composition of the otolith at the appropriate daily incre- 

 ment. Natural geochemical signatures in otoliths may 

 therefore be useful markers of environmental history 

 throughout the life of the individual and in turn, fish 

 stock composition (e.g., Campana et al., 1995). 



The use of geochemical signatures in otoliths as natu- 

 ral tags requires accurate and precise assays of otolith 

 composition. Electron probe micro-analysis (EPMA) has 

 been commonly used for probe-based analyses of otolith 

 chemistry (Gunn et al., 1992). However, detection lim- 

 its of approximately 100 ,ug/g limit the technique to a 

 relatively small number of minor elements in otoliths, 

 including Na, CI, K, and Sr (Campana et al., 1997). 

 Most of the elements measured by EPMA are probably 

 controlled by physiological rather than environmen- 

 tal factors, which may limit their usefulness in stock 

 identification studies (Campana, 1999). Nonetheless, a 

 number of researchers using EPMA have reported geo- 

 graphic differences in otolith chemistry (e.g., Thresher 

 et al., 1994). More recently, attention has focused on 

 inductively coupled plasma mass spectrometry (ICP-MS) 

 to assay elements that are typically below the detection 

 limits of EPMA. Laser ablation ICP-MS uses focused 

 Nd:YAG or excimer lasers to ablate specific locations on 

 the otolith. The vaporized material is then swept up by 

 a carrier gas into a plasma torch and analyzed by mass 

 spectrometry. Limits of detection of the technique are 

 typically on the order of 0.1-l^g/g, allowing for quan- 

 tification of several elements that cannot be assayed by 

 using EPMA including Mg, Mn, Ba, and Pb (Thorrold 

 et al., 1997; Thorrold and Shuttleworth, 2000). These 

 observations led Campana et al. (1997) to conclude that 

 EPMA and laser ablation ICP-MS were complementary 

 and that there is little overlap in the elements that are 

 accurately measured by the two techniques. Yet few 

 studies of otolith geochemistry have attempted to use 

 both approaches on the same samples. 



The objectives of this study are to determine if larval 

 walleye pollock from different geographic localities can 

 be distinguished based on elemental signatures in their 

 otoliths. By analyzing sagittal otoliths with both EPMA 

 and laser ablation ICP-MS, we hoped to identify greater 

 differences among locations than would have been pos- 

 sible by using either technique in isolation. If success- 

 ful, the study may provide a powerful tool for determin- 

 ing stock structure and tracing migration pathways of 

 walleye pollock in the north Pacific. These data could 

 then be used by managers of one of the world's largest 

 single species fisheries to direct the sustainable harvest 

 of this considerable natural resource. 



Materials and methods 



All fish used in the study were collected in the spring and 

 summer of 1999 from Alaska Fisheries Science Center 

 research cruises in the Bering Sea and Gulf of Alaska 

 (Fig. 1, Table 1). Fish of birth year 1999 were collected 

 within three months of spawning time to minimize the 

 likelihood of larval transport from other regions. In the 

 case of the Yakutat samples, fresh juvenile pollock were 

 removed from Pacific cod guts. Samples were collected 

 only when the pollock were readily identifiable and 

 heads were intact. Otoliths showed no visible sign of 

 degradation from digestive processes. Juvenile pollock 

 were frozen whole and transported to the laboratory for 

 otolith removal. 



Otoliths were removed from the fish and mounted on 

 petrographic slides in LR White resin (acrylic, hard- 

 grade). Larval otoliths were ground on one side to 

 expose the nucleus by using 500-grit paper and were 

 polished with 0.25-um grit diamond paste. Juvenile 

 otoliths were ground and polished in the sagittal plane 

 on both sides to maximize clarity of the nucleus during 

 microanalysis. 



Electron probe microanalysis 



After having been polished, the otoliths were cleaned 

 with Formula 409® and coated with a 30-nm layer 

 of carbon. They were subsequently analyzed with a 

 Cameca SX-50 electron microprobe equipped with four 

 wavelength dispersive spectrometers. A 15keV, 10 nA, 

 4-/jm diameter beam was used for all analyses. Counting 

 times, standards, detection limits, and analytical errors 

 are summarized in Table 2. Although Mg was analyzed 

 in all otoliths, in most cases it was below detection 

 limits and was therefore not used in the statistical 

 analysis. 



Laser ablation ICP-MS 



After having been ground and polished, otolith sections 

 were decontaminated before elemental analysis by using 

 laser ablation ICP-MS. Sections were rinsed in ultra- 

 pure water, scrubbed with a nylon brush in a solution 

 of ultrapure H,,0, triple rinsed with ultrapure 1%HN0 3 , 



