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



613 



Discussion 



Quantifying dispersal pathways of larval fishes in 

 marine environments is a difficult proposition. Marine 

 fishes typically produce on the order of lCH-lO 6 eggs in 

 a single spawning episode. These propagules are quickly 

 dispersed in large volumes of seawater, making recovery 

 of marked individuals difficult even if it were possible to 

 introduce an artificial tag into the larvae at the time of 

 spawning I Jones et al., 1999). Natural geochemical sig- 

 natures in otoliths offer a useful alternative to artificial 

 tagging approaches (Thorrold et al.. 2002). The technique 

 relies upon the assumption that larvae spawned at any 

 given location retain a unique elemental or isotopic sig- 

 nature in their otoliths that can be recovered some time 

 afterwards, and that variations in otolith geochemistry 

 are sufficient to distinguish among geographic locations 

 of interest. We found that elemental signatures in the 

 otoliths of larval walleye pollock differed significantly 

 geographically and with ontogeny. Samples at specific 

 points on the otoliths, at the nucleus, and shortly after 

 hatching, showed very similar patterns of variability, 

 suggesting that the technique will likely be a robust 

 method for identifying natal origins of walleye pollock 

 after suitable groundtruthing of known spawning loca- 

 tions (e.g., Thorrold et al., 2001; Gillanders, 2002). 



Elemental signatures in the otoliths of the larval 

 pollock were assayed by using EPMA and laser abla- 

 tion ICP-MS. Campana et al. (1997) noted that the two 

 techniques were largely complementary in terms of ele- 

 ments that could be reliably assayed, and indeed only 

 Sr was able to be quantified by both instruments in our 

 study. Patterns of geographic variability presumably 

 reflected the elements that were used in generating the 

 multivariate signatures produced by each instrument. 

 Elemental signatures from the EPMA data clustered 



into a Bering Sea group that included Shelikof Strait 

 but excluded Bristol Bay, a coastal grouping that in- 

 cluded Bristol Bay and Prince William Sound, whereas 

 samples from Yakutat were separated from all other 

 locations. Although sample sizes were smaller for the la- 

 ser ablation ICP-MS assays, the data identified a group- 

 ing of locations in multivariate space that included the 

 southeast Bering Sea, Prince William Sound, and Ya- 

 kutat, whereas samples from both the north Bering Sea 

 and Bristol Bay were separated from each other and the 

 other locations. The observation that locations did not 

 cluster into similar geographic groupings was probably 

 a function of different mechanisms of elemental incor- 

 poration in otoliths. Elements assayed by ICP-MS in the 

 present study substitute for Ca in the aragonitic lattice, 

 and are believed to primarily reflect ambient physico- 

 chemical differences among natal locations (Bath et al., 

 2000; Milton and Chenery, 2001; Bath Martin et al, 

 2003). However, with the exception of Sr, the elements 

 assayed by EPMA are likely under physiological regula- 

 tion and therefore probably do not directly reflect either 

 water chemistry or temperature (Campana, 1999). In 

 either case, the application of elemental signatures in 

 otoliths as natural tags of natal origins requires only 

 that the signatures allow accurate classification of the 

 natal origins of an unknown fish. A final caveat is nec- 

 essary because it remains possible that preservation 

 effects, particularly for labile elements that are not 

 incorporated into the aragonite lattice, may also have 

 contributed to at least some of the differences among 

 locations (Milton and Chenery, 1998; Proctor et al., 

 1998). If present, such effects would clearly confound 

 attempts to use elemental signatures as a natural tag 

 of natal origins (Thresher, 1999). 



It is important to note that although EPMA and laser 

 ablation ICP-MS provided complementary information 



