Thresher et al.: Otolith analysis of Nemadactylus macropterus 



829 



tions contribute significantly (P<0.001) to the dis- 

 crimination, is at the level of three groups and two 

 discriminant functions (i.e. Fig. 12); the third func- 

 tion, separating the two Tasmanian samples, is not 

 quite significant in the final step (P=0.057). Post-hoc 

 analyses (Steffe's P-test) of the discriminant func- 

 tions indicate that the samples from the Victorian 

 and NSW sites do not differ significantly and consis- 

 tently in any of the three functions, the GAB sample 

 differs from all other sites (which do not differ sig- 

 nificantly) in function 2, and none of the sites differs 

 significantly in function 3, though the western Tas- 

 manian sample nearly differs significantly from the 

 other sites. 



We draw three general conclusions from these 

 analyses. First, there are significant differences 

 among samples from different sites in terms of the 

 composition of the primordial region of their otoliths. 

 Second, analyses of the primordium itself, of a point 

 25 urn from the primordium, and of the mean value 

 for the region between 25 and 125 urn 

 from the primordium produce similar 

 results, indicating that distributional 

 differences of adults are manifest 

 through at least the first 125 urn of 

 otolith growth. And third, on the basis 

 of common patterns of composition, the 

 sites pool into three groups: one com- 

 posed of the NSW and the two Victo- 

 rian samples, a second consisting 

 uniquely of the Bight sample, and a 

 third consisting of the two Tasmanian 

 samples. In both cases where sites are 

 pooled, the pooled sites are geographi- 

 cally contiguous and nearest neighbors. 



Site-specific differences and 

 similarities in ontogenetic 

 variation in composition 



That differences in composition among 

 sites can be discerned at points as far 

 out as 125 urn from the primordium and 

 to an apparent age of 45-55 days post- 

 hatching suggest that delineation 

 among samples is not a function of con- 

 ditions specific to the spawning sites. 

 Any environmental differences must 

 encompass at least several weeks, per- 

 haps months, of larval development. To 

 assess the ontogenetic patterning of 

 these chemical differences, we com- 

 pared the concentrations of apparent 

 key elements for specimens from the 

 three pooled areas (NSW/Victoria, the 



ci 



Ca 



381 



1400 



Bight, and both Tasmanian sites) at several points 

 along their respective long growth axes. Five-point 

 filtered data were assessed at four positions: points 

 2-6, 6-10, 36^0, and 80-84. The first position is 

 immediately adjacent to the primordium; the second 

 immediately exterior to the first (and presumably 

 encompassing the second 2—3 months of planktonic 

 larval development); the third we estimate to corre- 

 spond approximately to the age when the prejuven- 

 iles recruit to the nursery areas; and the fourth, out- 

 ermost position, is the farthest along the growth axis 

 at which we had data for all specimens (the number 

 of points depended upon the length of the axis) and, 

 we estimate, corresponds to otolith deposition at an 

 age of 2-3 years. 



The results of the comparison (Fig. 14) lead to three 

 conclusions. First, the mean pattern of ontogenetic 

 change in composition is very similar for samples 

 from all three pooled sets of sites, e.g. Na and Sr 

 concentrations decline between points 2 and 6 and 



Pt 2 only 



Pts. 2-6 





: . ■° '.' 



CI 





Na 



.o • S * 

 s . 9 *° 



Sr 



3400 14 °0 



Figure 1 1 



Scattergrams of the relationships between Na and CI and between Sr 

 and Ca concentrations among the 64 adult Nemadactylus macropterus 

 examined, grouped by site, based on concentrations measured at point 

 2 only and the mean concentration for points 2-6, inclusive. Concen- 

 trations are reported in units of ppm by weight for all elements other 

 than Ca, which is reported as percent by weight. Sites are described 

 in Figure 1 and Table 1. Site key: E. Tas.=solid square; W. Tas.=solid 

 circle; NSW=open square; E. Vict.=open circle; GAB= plus sign; and 

 W. Vict.=open diamond. 



