606 



Fishery Bulletin 102(4) 



• ; "- 



North 



Bering 



Sea 



1$' 





■"\T7 



60N 



Bristol 15V"^C' 



ba y • j *?p* 



Prince 

 William 

 Sound 



-55N 



SE Bering 

 Sea 





*>•.. 



Shelikof 

 Strait 



^ A 



Yakutat 



•**>&* 



160W 



I 



Figure 1 



Locations of sampling sites for larval and juvenile walleye pollock (Theragra chalcogramma 

 the Gulf of Alaska and Bering Sea. 



) in 



sonified for 5 minutes in ultrapure H 9 0, and finally triple 

 rinsed again with Milli-Q water. The section was dried 

 under a positive flow hood for 24 hours and stored in a 

 polyethylene bag. 



Elemental analyses were conducted with a Finnigan 

 MAT Element2 magnetic sector field ICP-MS and Mer- 

 chantek EO LUV266X laser ablation system (Thorrold 

 and Shuttleworth, 2000). Instrument set-up was simi- 

 lar to that outlined by Giinther and Heinrich (1999). 

 An Ar gas stream was used to carry ablated material 

 from the laser cell to the ICP-MS. The carrier gas was 

 then mixed with the Ar sample gas and a wet aerosol 

 (1% HN03) in the concentric region of the quartz dual 

 inlet spray chamber. The wet aerosol was supplied by 

 a self-aspirating PFA micro-flow (20 /./L/min) nebulizer 

 attached to a CETAC ASX100 autosampler. Diameter 

 of the 266-nm laser beam was nominally 5 j.im, repeti- 

 tion rate was 5 Hz, and the scanning rate was set at 

 5 /im/sec. 



A typical run for an individual otolith consisted of a 

 blank sample (l%HNO :! only), a standard sample, five 

 laser samples, and then another blank and standard. 

 The number of laser samples in a run ranged from 5 

 to 15, depending upon the size of the otolith. All laser 

 runs began with a 70 fim x 70 /.im raster, centered on 

 the otolith nucleus. The laser software was then used 

 to trace out concentric lines, 720 /jm in length and ap- 

 proximately 40 /jm apart, which followed the contour 

 of individual growth increments from the raster to the 

 otolith edge. This approach produced reasonably high 

 spatial resolution (30-50 u,m) for life history scans 

 across otoliths while allowing sufficient acquisition time 

 to maintain measurement precision. 



We examined Mn/Ca, Sr/Ca, and Ba/Ca ratios in the 

 pollock otoliths by monitoring 48 Ca, 55 Mn, 86 Sr, and 

 138 Ba. Quantification followed the approach outlined 

 by Rosenthal et al. (1999) for precise element/Ca ra- 

 tios using sector field ICP-MS (Thorrold et al., 2001). 

 Quality control was maintained by assaying a dissolved 

 aragonite standard (Yoshinaga et al., 2000) every five 

 samples. The standard was introduced at the appropri- 

 ate time by moving the autosampler probe from the 

 solution containing the 1% HNO ;j to the standard solu- 

 tion, while maintaining the carrier gas flow through 

 the ablation cell. Elemental mass bias was calculated 

 by reference to known values of the standard, and a 

 correction factor was then interpolated and applied to 

 the laser samples bracketed between adjacent standard 

 measurements. Average u; = 40) within-run precisions 

 (RSD) of the standard measurements were all less than 

 1% (Mn/Ca: 0.16%, Sr/Ca: 0.16%, and Ba/Ca: 0.33%). 

 Long-term (5-month) estimates of the standard mea- 

 surements (n=40), again uncorrected for changes in 

 mass bias over time, were less precise (Mn/Ca: 5.6%, 

 Sr/Ca: 3.7%, and Ba/Ca: 5.6%). However, laser samples 

 were corrected for changes in mass bias by using the 

 laboratory standard. Precision of the technique was ap- 

 proximately 1% for all the ratios that we measured. 



Statistical analyses 



All elemental data were initially examined for nor- 

 mality and homogeneity of variance by using residual 

 analysis (Winer, 1971) and were found to conform to 

 the assumptions of ANOVA without the need for data 

 transformation. We therefore assumed that require- 



