152 



Fishery Bulletin 104(1) 



there was no evidence that size at the smolt stage of 

 postsmolts captured later in the season was larger than 

 those captured earlier in the season, indicating that 

 there was no size-selective attrition on smaller smolts 

 during August to October. The strength of these data is 

 that the samples were derived from collections of posts- 

 molts, not river returns or fishery catches, and thus it is 

 less biased and more representative of natural mortality 

 factors. It will take considerably more data to fully test 

 this hypothesis, but the highest priority should be given 

 to similar sample collections, especially those collected 

 earlier in the year (Holm et al., 2000). These data will 

 be pivotal in interpreting the relative impact of size 

 at ocean entry versus postsmolt growth for determing 

 postsmolt survival. 



Understanding survival during the postsmolt period 

 is of great importance in the face of declining stock 

 abundance and possible stock extirpations over por- 

 tions of the range of Atlantic salmon (Anderson et al., 

 2000). With evidence showing the success of conserva- 

 tion efforts to increase freshwater populations (Swans- 

 burg et al., 2002), the range of potential life stages 



0.12- 



0,10' 



0.08- 



0.06  



0.04 



0.12 



O 0.10 



0.08- 



0.06 



0.04 



1982 

 1983 

 1984 

 1982 

 1983 

 1984 



B 



„«oe 



o**.>-?' 



^^^■■^ 



if-- 



0.01 



0.05 



0.02 0.03 0.04 



Circuli spacing, late freshwater period (mm) 



Figure 3 



Relationships between circuli spacing during the early marine 

 period (A) and circuli spacing during the summer marine period 

 (B) and circuli spacing during the late freshwater period for 

 Atlantic salmon (Salmo salar) in the Gulf of St. Lawrence. 



and habitats causing recruitment failure is reduced. 

 Evidence at the population level showing coherence be- 

 tween early marine growth and survival at sea is still 

 a long way from providing sufficient evidence to fully 

 describe survival mechanisms or climate linkages. The 

 hypothesized climate forcing appears to be related to 

 a mismatch between ocean entry time for smolts and 

 the thermal regime they find during their transition 

 to marine life (Friedland et al., 2003b). The concept is 

 supported by the growth response of postsmolts over a 

 range of temperatures that they would likely encounter 

 during that period of time. Atlantic salmon postsmolts 

 follow a nonlinear growth response to temperature 

 and optimum growth occurs at 13°C (Handeland et al., 

 2003). Variations in migration timing could result in 

 fish encountering ocean conditions that are too warm 

 or cool; these variations could explain the different 

 associations of North American and European stocks 

 to ocean temperatures. It would also be interesting to 

 know if the thermal growth optimum is robust to vary- 

 ing feeding rations, which may be an issue for regional 

 stock groups in the North Atlantic. 



The transition period from freshwater to ma- 

 rine life is characterized by a number of dy- 

 namic changes in growth and predation. The 

 highest mortality rates for postsmolts occur 

 during the first weeks at sea (Eriksson, 1994). 

 As a consequence, even a minimal variation in 

 these rates will have a large impact on adult 

 recruitment. Size-selective mortality on juve- 

 niles has been demonstrated for other salmo- 

 nids such as sockeye salmon {Ojjcorhynchiis 

 nerka) and has been shown to be seasonally 

 concentrated (West and Larkin, 1987). Size- 

 specific predation often occurs but predators, 

 such as adult bird species, are not likely to 

 grow at corresponding rates to those of the fish 

 (Dieperink et al., 2002). Thus, size at ocean 

 entry and the ability to grow out of the size 

 range vulnerable to specific predators must 

 affect the dynamics of predation rate. The pe- 

 riod of predation vulnerability is mediated by 

 growth rate; therefore the argument returns 

 to what confers faster growth on postsmolts in 

 some years? 



One dynamic that challenges all stocks is the 

 transition from predominantly invertebrate to 

 piscivorous prey. Some authors suggest it is the 

 success in making this transition, or the time 

 it takes to make the transition, that dictates 

 the amount of time the postsmolts take to grow 

 out of the predation-vulnerable size range (Du- 

 til and Coutu, 1988; Andreassen et al., 2001; 

 Salminen et al., 2001). Depending on prevail- 

 ing foraging conditions, this transition may be 

 promoted by increased smolt size, as was the 

 case in the northern Baltic Sea in 1990-93 

 (Salminen et al., 2001). One potential expla- 

 nation for the contradictory views on the rela- 

 tionship between pre- and postsmolt growth in 



