150 



Fishery Bulletin 104(1) 



marine growth in Atlantic salmon, whereas Lundquist 

 et al. (1988) and Salminen (1997) reported positive 

 relationships. Einum et al. (2002) reported an inverse 

 relationship between pre- and postsmolt growth of At- 

 lantic salmon and suggested that phenotypic charac- 

 teristics favoring growth in one environment may not 

 necessarily favor growth in another. The body of mixed 

 results sheds little light on whether smolt size confers a 

 growth advantage to postsmolts. It also leaves in doubt 

 the importance of freshwater experience on postsmolt 

 survival. Common to this body of work is the depen- 

 dence on samples coming from river returns and fishery 

 catches that may not be representative of the full range 

 of growth signatures in postsmolt populations because 

 they do not include data from mortalities and often do 

 not include data from some return age groups. 



In this study we report on an analysis of scale growth 

 indices from a collection of postsmolts from the Gulf of 

 St. Lawrence. We measured freshwater growth signa- 

 tures representative of smolt size and freshwater growth 

 prior to migration and postsmolt growth was partitioned 

 by season. These are samples of a life stage in an in- 

 termediate phase of marine life that is infrequently 

 sampled and may be free of some of the assumed biases 

 associated with samples from spawning fish. 



Figure 1 



Postsmolt scale of Atlantic salmon iSalmo 

 aalar) in the Gulf of St. Lawrence show- 

 ing focus, freshwater growth zone, and 

 postsmolt growth zone. 



Material and methods 



We collected data on scale circuli spacing that was rep- 

 resentative of the freshwater and postsmolt growth for 

 juvenile salmon captured in the Gulf of St. Lawrence. 

 These postsmolt salmon were collected in 1982-84 and 

 were originally reported in Dutil and Coutu (1988). 

 They were captured in experimental gill nets along 

 the northwest shore of the Gulf during the months of 

 August to October. 



Freshwater and postsmolt growth descriptors were in- 

 terpreted from circuli spacing patterns deposited on the 

 scale. Scales were cleaned and mounted between glass 

 slides and the spacings of scale circuli were measured 

 with a Bioscan Optimas (Media Cybernatics, Inc., Sil- 

 ver Spring, MD) image processing system. Freshwater 

 zone length (FZL) was taken as the total distance from 

 the center of the scale focus to the transition between 

 freshwater and postsmolt growth (Fig. 1). This tran- 

 sition zone is defined by the appearance of the first 

 marine intercirculus spacing, i.e., as a wider spacing 

 (according to the reader's judgment) than the progres- 

 sion of spacings in the freshwater zone. The FZL ends 

 at the last freshwater circulus. FZL was interpreted as 

 a proxy for smolt length at migration. The mean of the 

 last five circuli spacings during the freshwater phase 

 was computed for each sample (CSLF, circuli spacing 

 during last freshwater period). This circuli spacing 

 index was interpreted as an indication of freshwater 

 finishing growth (the final phases of freshwater growth 

 prior to migration). Two circuli spacing indices were 

 extracted from the postsmolt zone, the mean spacing 

 between intercirculi 2 through 6 — an interval which 



was interpreted as the growth index for the early ma- 

 rine period (CSFM, circuli spacing during first marine 

 period) — and the mean intercirculi spacing of the next 

 five pairs occurring later in the postsmolt growth — an 

 interval interpreted as an index of summer growth 

 (CSSM, circuli spacing during summer marine period). 

 The total length of the postsmolt growth zone was also 

 extracted (MZL, marine zone length), which was the 

 distance from the freshwater-marine transition, starting 

 at the last freshwater circulus to the outer edge of the 

 scale. All measurements were made on a single scale 

 from each specimen along the 360° axis of the scale. 



We tested the functional relationship between fresh- 

 water and marine growth, early and late marine 

 growth, and seasonal progression of the growth zones 

 using linear regression. We tested the significance of 

 the slope parameters for each relationship as an indica- 

 tion of variable dependency. 



We also considered whether these data might contrib- 

 ute to our understanding of size selective mortality by 

 constructing a time series plot of FZL and MZL against 

 date of capture. An anticipated positive trend in MZL 

 would reflect growth during the marine phase, but any 

 trend in FZL would reflect size-specific removals at- 

 tributable to size at ocean entry. 



Results 



Measurements were taken from the scales of 587 posts- 

 molts collected during 1982-84. FZL averaged 0.67 mm 

 (SD = 0.163) for all samples. CSLF averaged 0.022 mm 

 (SD = 0.0049), reflecting the slower growth and more 

 narrowly spaced circuli of the freshwater zone. The two 



