Thresher et al.: Otolith analysis of Nemadactylus macropterus 



825 



#304 



#312 



#339 



6000 



E 



Q. 



C 



o 



c 



Q) 



o 

 c 

 o 

 O 



140 



140 



Point number 



Figure 7 



Comparisons of life history transects of the left and right sagittae of three adult Nemadactylus macropterus 

 based on unfiltered data for four representative elements (Sr, Na, S, and CD. The dashed horizontal line in 

 the CI plots is the minimum detection limit for the element. The comparison is based on identical analyti- 

 cal conditions and an attempt to track as closely as possible the same growth axis in the two otoliths in 

 each pair. 



ences between otolith pairs are real, appear to be 

 more common in some elements than in others (e.g. 

 evident in CI, but not in Sr), and may be more com- 

 mon near the otolith margin than closer to the pri- 

 mordium. The data are limited but clearly indicate 

 that slight differences in elemental concentrations, 

 particularly CI and S, should be used with caution 

 for stock delineation. 



Evaluation of stock structure 



We hypothesized that stocks of N. macropterus would 

 differ in either or both spawning grounds or times 

 and that these differences would result in diagnos- 

 tic patterns of composition in the first forming part 

 of the otolith. However, in practice, compositional 

 data for the primordium itself proved of low quality 



because of the specimen preparation required for 

 EPMA and the incremental structure of otoliths. In 

 most specimens, minute cracks or a pit several mi- 

 crons in diameter developed at the primordium dur- 

 ing preparation, the latter because of the 'plucking' 

 of the primordium from the otolith center during 

 polishing. Our previous work (Gunn et al., 1992) in- 

 dicated that topographic irregularities degrade 

 EPMA data because of unpredictable patterns of x- 

 ray absorption. A comparison of data for point 1 (on 

 and immediately around the primordium) and point 

 2 (25 urn from the primordium) for the adults sup- 

 ports this conclusion: for all elements except CI the 

 variance in estimated concentrations is 36 (Na) to 

 270 % (S) higher for point 1 than for point 2 (Fig. 8). 

 Nevertheless, for all elements, concentrations at 

 point 1 are significantly correlated with those at point 



