Thresher et al.: Otolith analysis of Nemadactylus macropterus 



831 



signature for each nursery area (based on otolith 

 material deposited during residence in the nursery 

 areas) and using these to classify adults collected in 

 different regions. In essence, the former assesses how 

 juveniles from each of the putative populations are 

 distributed among nursery areas, whereas the lat- 

 ter assesses the contribution of each nursery area to 

 adults collected at each site. With specific regard to 

 N. macropterus, if SE Tasmania is the sole nursery 

 area for the species in Australia, then we would ex- 

 pect that 1) the complete range of chemical patterns 

 documented in the adults, at all sites sampled, would 

 be seen in juveniles collected in the single, common 

 nursery area, and 2) adults collected at all sample 

 sites would have a nursery area 'fingerprint' similar 

 to that of the juveniles collected. 



TAS. 



NSW/VICT. GAB 



3600 



2800 

 3000 



Na 



c 

 o 



c 



Q) 

 U 



C 

 O 



O 



1600 

 42 



(%) 



34 

 800 



Sr 



Ca 



CI 



U 



2 6 36 80 2 6 36 80 2 6 36 80 



Point analyzed 



For the first analysis, the three-site discriminant 

 functions developed from the adults were used as a 

 training set to classify each of 116 recently settled 

 N. macropterus. The data for the juveniles were ac- 

 quired in the same way as for the adults. Analysis is 

 based on the mean values for points 2-6 from the 

 primordium. Most juveniles were collected in SE 

 Tasmania; a small number were also collected at 

 Phillip Island, Victoria (Table 1). 



Most juveniles examined fell within or close to the 

 areas in discriminant function space defined by the 

 adult groups (Fig. 15); only one, with an exception- 

 ally high value on the function 2 axis, did not match 

 the characteristics of at least one of the three adult 

 groups. Moreover, most juveniles classified with the 

 adults collected in the same area. Of the 106 juveniles 

 caught in Tasmania, all but 25 classified with the Tas- 

 manian adult samples, and of these, 13 classified am- 

 biguously, with a probability >25% of being Tasmanian. 

 Overall, only 7% of the Tasmanian-collected juveniles 

 had a probability of <10% of classifying with the Tas- 

 manian-caught adults (Fig. 16). Samples from the five 

 Tasmanian sites were distributed similarly in two-func- 

 tion space (Fig. 15), though the variance was conspicu- 

 ously higher at one site (Cygnet). 



The pattern was similar for juveniles collected off 

 Victoria (Phillip Island), although sample sizes were 

 too small to draw strong inferences. Of the 10 indi- 

 viduals examined, six classified with the NSW/Vic- 

 torian adults, three classified with the Tasmanian- 

 caught adults (at probabilities ranging from 72 to 

 85%), and one classified with the Bight-caught adults 

 (atP=63%). The probability that the Victorian juve- 

 niles classify with the NSW/Victorian adults is mark- 

 edly bimodal (Fig. 16) with peaks at >95% and be- 

 tween 5-10%. That is, most individuals had either a 

 very high or very low probability of classifying with 

 the local adults. A similar pattern may also be the 



Figure 14 



Comparisons of mean concentrations of four elements (Ca, 

 Na, Sr, and CI ) at four points along the life history scan of 

 each adult Nemadactylus macropterus pooled by groups 

 identified by linear discriminant function analysis ( LDFA). 

 For each individual and point, data were calculated as the 

 mean of the 5-point moving average, beginning at the point 

 indicated (i.e. 2=mean of points 2-6, inclusive; 6=mean of 

 points 6-10, inclusive, etc.). K and S are not depicted as 

 the former generally varies similarly to Na, whereas the 

 latter did not differ significantly among groups. Horizon- 

 tal dashed line in CI plot indicates minimum detection 

 limit. Vertical lines about each mean indicate one stan- 

 dard error of the mean. Concentrations are reported in 

 units of ppm by weight for all elements other than Ca, 

 which is reported as percent by weight. 



