DUTIL AND COUTU: EARLY LIFE OF ATLANTIC SALMON 



1985 to get hourly records of temperature near 

 the surface, but portions of these records are miss- 

 ing. 



Fish were preserved at -15°C for less than 

 4 months in 1982 and less than 2 months in 1983, 

 1984, and 1985. Fork length was measured to 

 nearest mm and weight to nearest g. Condition 

 index values were calculated as 



C/ = sin 1 



100  W 



0.5 



CI = condition index; W = weight (g); L = length 

 (mm); s = slope of the length- weight relationship 

 for age groups combined. Scales, usually taken 

 below the dorsal fin close to the lateral line, were 

 cleaned and mounted on glass slides. Stomachs 

 were preserved in formalin (10%) or alcohol (70%) 

 to be analyzed later for their contents. The sex 

 was determined visually and checked histologi- 

 cally on a subsample of postsmolts collected in 

 1982. Gonads were weighed to nearest mg and 

 only in 1982. Gonadosomatic index (GSI) was cal- 

 culated as 



GSI = 



GW 



TW-GW 



100 



where GW = gonad weight; TW - total weight. 



Scales showing no focal regeneration were ex- 

 amined for age determination by 2-4 readers. 

 Readings were discussed and a consensus was 

 reached in most cases. Reported ages are smolt 

 ages plus 2-4 months. Fork length at smoltifica- 

 tion (mm) was back-calculated as 



LS = 



(LC - 34)  RS 

 RC 



+ 34 



where LS = back-calculated smolt length; LC 

 = body length of post smolt; RS = scale radius at 

 smolt check; RC = scale radius of postsmolt. 

 Length at scale formation is assumed to be 34 

 mm. Scales were also examined for any mark in 

 the postsmolt zone that could be of potential in- 

 terest. Hence the number of individuals showing 

 a transition zone between the riverine and 

 marine increments was determined in 1982, and 

 the number of individuals showing a summer 

 check was determined in 1982, 1983, 1984, and 

 1985. Scale radii were measured along the 

 postero-anterior axis on 4 scales per individual in 

 1982, and on 1 scale per individual in 1983, 1984, 

 and 1985. Projected scales were measured at con- 



stant magnification on a digitizing pad connected 

 to a personal computer. Fork length at formation 

 of a summer check (mm) was back-calculated for 

 1982 and 1983 as 



LF = 



(LC - 34) • RM 

 RC 



+ 34 



where LF = back-calculated length; LC = body 

 length of postsmolt; RM = scale radius at summer 

 check; RC = scale radius of postsmolt. 



Stomach contents were examined in detail in 

 1982. Results are expressed as percentage of oc- 

 currence (number of stomachs containing an item 

 divided by the number of stomachs examined) 

 and percentage by number (the count number of 

 an item divided by the count number of all items). 

 Food items that could not be identified are re- 

 ferred to as remains. In as far as their condition 

 allowed, prey fish were usually measured to the 

 nearest cm. Stomach contents examined in 1983, 

 1984, and 1985 were consistent with the 1982 

 conclusions, so the results are not presented. 



Results 

 The Age and Sex Composition 



Three-year-old female salmon dominated the 

 catch both in 1982 and 1983 (Table 1). Postsmolts 

 ranged from 2 to 5 years in 1982 and from 2 to 4 

 years in 1983. Five-year-old salmon made up less 

 than 4% of the catch in 1982. Three-year-old and, 

 to a lesser extent, 4-yr-old salmon dominated the 

 catch both in 1982 and 1983 (82% of the catch); 

 mean smolt age was 3.2 ± 0.7 years. Age composi- 

 tion in 1983 differed from 1982 because of an in- 

 crease in the percentage of 2-yr-old salmon (12% 

 to 21%) and a decrease in the percentage of 5-yr- 

 old salmon (x^ = 15.96, P < 0.01). The sex ratio 

 (male:female) for 1982 and 1983 combined was 



Table 1 . — The age and sex composition of the catch of postsmolt 

 Atlantic salmon along the north shore of the Gulf of St. Lawrence, 

 1982 and 1983. 



199 



