influenced by oceanic conditions, the high seas 

 fishery, and the amount of time the salmon spend 

 in the ocean — factors that may mask the effects 

 of the size of smolts on mortality. Estimates of 

 ocean mortality of Bristol Bay sockeye salmon 

 are available for only the Naknek and Ugashik 

 systems (table 30) . Age I smolts in the seaward 

 migration from the Naknek system in 1958 were 

 smaller (91 mm.) than smolts of the same age in 

 1959 (97 mm.) and 1960 (jy mm.) . Yet the 1958 

 group suffered the lowest total ocean mortality 

 (from smolt to return — table 30) . Age I smolts 

 of the same year class also had the lowest total 

 ocean mortality for the Ugashik system. The age 



I smolts in the 1958 Ugashik migration were 

 larger, however, than those of 1959 and 1960. 

 Any effect that smolt size in 1958 might have had 

 on ocean mortality may have been masked by 

 effects of environmental conditions especially 

 conducive to good survival. Valid comparisons 

 can be made only if groups of smolts of different 

 sizes encounter conditions that are similar. Ob- 

 viously, this situation is rare. 



The data in table 30 afford an opportunity for 

 comparison of ocean mortality for age I and age 



II smolts. Age II smolts without exception suf- 

 fered a lower total mortality. How much of this 

 difference can be attributed to their larger size 

 rather than time of entry into the estuary cannot 

 be determined, for age II smolts generally enter 

 the sea at least 1 or 2 weeks before age I smolts. 



RELATION BETWEEN FRESH-WATER AGE OF 

 SMOLTS AND NUMBER OF ADULTS IN 

 PARENT ESCAPEMENT 



The tendency for slow-growing sockeye salm- 

 on juveniles to remain an additional year in 

 fresh water before they migrate seaward has 

 been documented (Foerster, 1937; Barnaby, 

 1944; Krogius and Krokhin, 1956; Koo, 1962). 

 Data presented by Foerster (1944) suggest that 

 in some studies, however, the tendency is re- 

 versed, and slow-growing sockeye salmon go to 

 sea at age I. Also, an increase of a year or more 

 in length of residence in fresh water accom- 

 panied an increase in growth rate that resulted 

 from reduction of populations (Krogius, 1961). 

 Little is known regarding the factors that cause 

 juvenile sockeye salmon to migrate to sea. 



Wood System 



We might anticipate in this system a higher 

 percentage of age II smolts from the larger 



Table 30. — Ocean mortality of sockeye salmon by fresh-water 

 age, Naknek and Ugashik systems 



Millions Mm 



Naknek system 

 1958 



Age I 



Age II 



Total- _. 



1959 



Age I 



Age II 



Total. __ 



1960 



Age I 



Age II 



Total _ - . 



1961 



Age I 



Age II 



Total... 



Ugashik system 

 1958 



Age I. 11 438 93 



Age II 222 112 



Total.. 



1959 



Age I 



Age II... 



Total . . 



1960 



Age I 



Age II... 



Total.. 



1961 



Age I 



Age II... 



Total 3 802 



' The 1960 totals include age HI smolts; 

 years. 



Years 



2.17 

 2.04 



Percent 



69.7 

 49 7 



none were recorded in other 



spawning populations than from the smaller be- 

 cause of the inverse relation between number of 

 parents and growth of progeny during their ini- 

 tial year of lake residence. This tendency is sug- 

 gested by the data (fig. 20) , but the progeny of 

 the large escapement year, 1959, are an out- 

 standing exception to the theory that slowly 



446 



U.S. FISH AND WILDLIFE SERVICE 



