Friedland et al.: Growth patterns in postsmolts of Salmo solar 



475 



comparison stocks. The data for the com- 

 parison stocks were grouped by sea-age; 

 thus, the postsmolt scale data were com- 

 pared with a total of five scale growth sig- 

 nals. The statistical comparison was sim- 

 plified to a two-group ANOVA, namely to 

 comparing circuli spacing of the first ten 

 circuli pairs of the postsmolts with a com- 

 bined sample of all the comparison stocks. 



Results 



Circuli spacing patterns for the postsmolt 

 growth zones of Gulf of St. Lawrence 

 postsmolts were significantly different for 

 the three year classes we examined. Cir- 

 culi spacing for the first few pairs, associ- 

 ated with the first few weeks in the ma- 

 rine environment, were similar in all three 

 years, measuring approximately 0.050 to 

 0.055 mm (Fig. 3). However, beyond these 

 initial pairs the spacings diverged into 

 three distinct patterns. Spacings for fish 

 captured in 1982, based on the mean spacings for 

 pairs 4-8, were the widest, generally larger than 

 0.065 mm. Circuli spacings in the same region of the 

 scale were progressively smaller for samples collected 

 in 1983 and 1984. In 1983, most of the mean spac- 

 ings did not exceed 0.060 mm and for 1984 they did 

 not exceed 0.055 mm. ANOVA results show that the 

 yearly spacings means were significantly different 

 for all but one of the first ten spacing pairs (Table 2). 

 The means for spacing pair number 2 is not signifi- 

 cantly different because the rank order of the means 

 changes between spacing pairs 1 and 3. 



Circuli pair 



Figure 3 



Circuli spacing versus circuli pair for Gulf of St. Lawrence postsmolts 

 for three smolt year classes, 1982-84. Error bars mark 95'7f confidence 

 intervals. 



Within-year comparisons of Gulf of St. Lawrence 

 samples poststratified by collection date suggest that 

 in one year the same group or population offish were 

 sampled during the entire growth season; whereas, 

 in another year, the postsmolt assemblage appeared 

 to change during the sampling period. All three 

 poststratified groups formed from the 1982 data 

 showed a similar pattern of circuli spacing versus 

 circuli pair through pair number 12 (Fig. 4A). After 

 circuli pair 12, the mean spacing for the three 

 poststratified groups began to diverge; however, the 

 mean spacing for these circuli pairs was estimated 



