47 o Memoir Sears Foundation for Marine Research 



the surface, and go or are carried downstream {64)- While the resulting downstream 

 movement is facilitated by freshet conditions and although the peak of the run is usually 

 associated with a freshet, there is evidence to suggest that some other factor may be 

 responsible for the initiation of the migration. As long ago as 1926, Bull produced evi- 

 dence suggesting that floods did not provide the departure stimulus (25). However, his 

 suggestion that the initiation of the run was not associated with rainfall has not been 

 accepted. Swain has recently concluded that the immediate stimulus is either a rise in 

 temperature or some factor associated with a rise in temperature {125). Huntsman has 

 suggested that freshets appear to affect migration not only through carrying fish down- 

 stream but through breaking up their "homes" {6y: 399; JO'. 257). 



Normally the main smolt migration occurs in the spring or early summer, chiefly 

 at night, with minor seaward movements at other times of the year. In the rivers of the 

 Maritime Provinces of Canada, the main migration takes place in May and June. In 

 the Little Codroy River, Newfoundland, the run during 1955 took place from May 1 1 

 to July 22, with the peak during June i — 15 (5: 63, 64). In the northerly part of their 

 range, in rivers flowing into Ungava Bay, most of them migrate in June and July, 

 though smolts may be taken there at all seasons {ll5)- It is evident from the foregoing 

 that all members of a year-class do not migrate in the same year, nor at the same time 

 in any year. 



Life at Sea. After moving into salt water, the smolts apparently spend some time 

 in the estuaries and river mouths and descend to deeper water with the onset of winter 

 (18). The length of time they remain at sea before returning to fresh water on their first 

 spawning migration varies from one to four years, and in some areas up to five years. 

 As in river life, there is variation in growth and size at sea brought about by various 

 factors. Thus fish of the same age from different rivers often grow at different rates 

 while at sea, and members of one year-class may reach weights that are different from 

 those of another year-class from the same river with the same length of sea-life. Gen- 

 erally, the longer they continue to feed at sea, the larger they are when they return, 

 growing more rapidly in summer than in winter. Those that spend two or more con- 

 tinuous years at sea before spawning will in most instances be larger fish than those 

 of a similar age from the same river that have survived the ordeal of spawning, since 

 the growth of the latter has been interrupted by fasting for a longer or shorter time, 

 depending on how long they have been in fresh water before spawning. 



Salmon develop rapidly during their sea life. Thus smolts may grow from only 

 one or two ounces to as much as six or seven pounds and a length of 16 inches or 

 more in their first year at sea. 



There is comparatively little certain knowledge about the habits of Atlantic Salmon 

 when they are in salt water. Most of what we know about their movements in the 

 sea is derived from data on fish caught in nets set by commercial fishermen. Since such 

 nets are fished only in certain restricted areas, there is little knowledge of this species in 

 places where no fishing has been carried on. We have no knowledge, for instance, as 

 to whether or not they are to be found in midocean, far from any shore, because there 



