826 



Fishery Bulletin 92(4), 1994 



4000 



2000 

 C\j 2000 



c 

 o 

 Q_ 



42 



32 



32 



.*#, 

 '<}*' 



Na 

 r 2 = 0.21 



1000 



r= 0.09 





4000 1 000 

 ,1000, 



200' 



4000 200 



1000 







Ca 

 r 2 = 0.18 





r = 0.25 



3001 



42 300 



1200 



1000 



1200 



Point 1 



Figure 8 



Scattergrams of the relationship between concentrations of the six detected elements (Ca, 

 Na, Sr, K, S, and CI) as measured on the otolith primordium (point 1) and as measured at a 

 point 25 pm from the primordium (point 2) along the long growth axis of 64 adult 

 Nemadactylus macropterus. Concentrations are reported in units of ppm by weight for all 

 elements other than Ca, which is reported as percent by weight. Correlations between the 

 two estimated concentrations are significant for all elements. The slope of the regressions 

 differ significantly from 1.0 for all elements except CI, although for several elements (Na, K, 

 and Ca) the trend line is very close to a slope of 1 and the calculated regression shifted 

 mainly because of relatively few very low values at point 1. The latter are expected conse- 

 quences of pitting and cracking on the primordium. Dashed lines in the CI scattergram 

 indicate minimum detection limits. 



2 (r 2 values from 0.09 (S) to 0.85 (CD). Slopes for re- 

 gressions between point 1 and point 2 data are typi- 

 cally not equal to 1, which for most elements prob- 

 ably reflects the highly proteinaceous nature of the 

 primordium. Nonetheless, we conclude that differ- 

 ences among specimens evident in the primordium 

 are also evident adjacent to the primordium, where 

 they can be measured more precisely. 



The distributions of mean concentrations of the six 

 elements for the 68 adult N. macropterus analyzed 

 are depicted in Figure 9. Three of the six (K, S, and 

 CD are significantly skewed to higher concentrations. 

 Ca also shows evidence of a weak skew, and Sr evi- 

 dence of a weak bimodality. Mean concentrations of 

 four of the six elements differ significantly among 

 sites (Fig. 10), the exceptions being Sr and S. The 

 differences are manifest in both the point 2 and point 

 2-6 data and are of similar pattern and comparable 

 magnitude in both data sets. For most elements, er- 



ror bars are smaller in the filtered data, which pre- 

 sumably reflects the reduced effect of random mea- 

 surement errors. Differences among sites are great- 

 est for CI: the mean values for three sites (eastern 

 and western Tasmania and the Great Australian 

 Bight) do not differ significantly from the minimum 

 detection limit (MDL), whereas means for the Victo- 

 rian and New South Wales (NSW) samples are well 

 above the MDL and do not differ significantly from 

 each other. There are suggestions of a similar, though 

 less pronounced grouping of sites in Na and K (con- 

 centrations in the Victorian and NSW fish higher 

 than in those from Tasmania and the Bight) and Ca 

 ( lower concentrations in Victorian and NSW adults ). 

 The grouping of sites was examined further by plot- 

 ting Na/Cl and Sr/Ca ratios for specimens from each 

 of the six sites (Fig. 11 ). These ratios were chosen on 

 the basis of a preliminary survey of the data as likely 

 to separate sites. As expected, the scatter of points 



