508 



Fishery Bulletin 104(4) 



The second inference from our results is that region- 

 al patterns in otolith shape are not always consistent 

 among cohorts and may be subject to temporally dy- 

 namic events such as large-scale environmental per- 



C/5 



tl 



-0.4 - 



0-8 



06 



0.4 



0.2  



0.0 



-0.2 



■0.4 



-0.6 



B 



Lizard Island Townsville Mackay Storm Cay 



Figure 5 



Mean ±SE of discriminant function (DF) I scores from otolith 

 morphological variables (Fourier harmonics and shape variables 

 combinedl per region for lA) cohort 1995, and (B) cohort 1999 

 in four-year-old common coral trout iPlectropomux leopardus) 

 from the Great Barrier Reef. (The data were pooled across reefs 

 within regions.) 



turbations. Ignoring such temporal signals could give 

 misleading information about stock structure. A single 

 homogenous stock of P. leopardiis may have been sug- 

 gested if otolith structure based on the 1999 PCA re- 

 sults alone were considered, whereas two poten- 

 tial stocks would be inferred from both the PCA 

 and CDA analyses of the 1995 data (Figs. 2A; 

 5A). In addition, cohorts were distinguishable 

 within all regions based on the frequency dis- 

 tributions of DF scores. Other investigations of 

 the interannual stability in otolith morphology 

 of marine fishes have found differences between 

 years, as well as age groups (Castonguay et al., 

 1991; Campana and Casselman, 1993; Begg and 

 Brown, 2000; Begg et al., 2001). Consequently, 

 when using otolith morphology to investigate 

 stock structure, it is preferable to compare mul- 

 tiple cohorts with individuals of the same age, 

 collected over several years. Such a sampling 

 design would minimize confounding spatial 

 variation in otolith shape with particular times 

 of sampling and derive a time-averaged assess- 

 ment of the spatial structure of a stock. 



The spatial and temporal patterns in otolith 

 shape are in part consistent with the assess- 

 ment (based on otolith microchemistry) of stock 

 structure of P. leopardus on the Great Barrier 

 Reef (Bergenius et al., 2005). Differences in 

 otolith chemistry of P. leopardus collected at 

 the same spatial and temporal scales as those 

 considered in this study have indicated the 

 presence of two or three regional stocks. Thus, 

 although the number and potential boundaries 

 of phenotypic stocks based on otolith chemistry 

 and otolith structure remain uncertain, the 

 combined results of these studies emphasize 

 the potential presence of several stocks of P. 

 leopardus on the GBR and indicate a north- 

 south demarcation. Differences in both otolith 



