FISHERY BULLETIN: VOL. 85, NO. 3 



Inspection of the diversion rates over the last 

 33 years suggested that after 1977 a change oc- 

 curred in the migratory patterns of Fraser River 

 sockeye approaching the coast. During the period 

 1978-85, unprecedented rates of 70 to 80% have 

 used the northern route (Fig. 1). This change in 

 migratory behavior coincided with a remarkable 

 prolonged warming period in the northeast 

 Pacific Ocean (Chelton 1984; McClean 1984). The 

 trend culminated in the extended warm-water 

 anomaly of 1983 along the coast of British Colum- 

 bia, which was associated with the 1982-83 El 

 Nino event that occurred in the equatorial Pacific 

 Ocean. This event was one of the most extreme of 

 the century (Mysak 1985). We therefore carried 

 out separate analyses for the two periods, 1953-77 

 and 1978-85. 



The regression analyses for the 1953-77 period 

 identified Fraser River discharge as the only sig- 

 nificant iP < 0.01) factor, explaining 45% of the 

 variance (Table 2; Fig. 7B). This positive relation- 

 ship between northern diversion rates and Fraser 

 River discharge was also suggested by Wickett 

 (1977) for the same time period. 



For the period 1978-85, the regression analyses 

 indicated that SST at Kains Island was the only 

 helpful predictor, explaining 85% of the variance 

 (Table 2; Fig. 7C). A strong positive relationship 

 between northern diversion rates of Fraser River 

 sockeye salmon and SST at Kains Island was also 

 noted for the years 1973-83 by staff of the Pacific 

 Salmon Commission (IPSFC 1984). 



DISCUSSION 



We suggest that sockeye salmon returning to 

 the Fraser River may have been influenced by 

 year-to-year changes in ocean conditions during 

 and between the periods 1953-77 and 1978-85. 

 The relationships of sockeye migration to sea sur- 

 face temperature and river discharge will be dis- 

 cussed separately. 



Sea Surface Temperature and 

 Sockeye Salmon Migration 



Leggett's (1977) review of fish migration con- 

 cluded that oceanic fish migrations largely repre- 

 sent the continuous optimization of physiologi- 

 cally important conditions. Temperature is an 

 oceanographic feature whose importance in fish 

 physiology is well established (Brett 1970). While 

 evidence indicates that thermal conditions may 

 be correlated with the timing of salmon migra- 



tions (Burgner 1980; Blackbourn in press) or the 

 route of their return migration to coastal waters 

 (this study), it is not clear how temperature af- 

 fects salmon behavior. Temperature might di- 

 rectly influence salmon in some way or it might 

 merely correlate with some other oceanographic 

 feature influencing them such as eddies and cur- 

 rents (Mysak 1986), or the abundance or species 

 composition of prey items (Fulton and LeBrasseur 

 1985). If so, a correlation of salmon behavior with 

 temperature could mislead attempts to under- 

 stand the control of migration. 



Alternatively, temperature may indeed have a 

 direct impact on sockeye salmon. There is consid- 

 erable evidence that temperature is correlated 

 with the distribution of marine fishes (Brett 1970; 

 Laurs and Lynn 1977; Laurs et al. 1977; Magnu- 

 son et al. 1980). Manzer et al. (1965) and French 

 et al. (1976) summarized the distribution of sock- 

 eye salmon in relation to sea surface tempera- 

 ture. While waters of certain temperatures were 

 generally devoid of sockeye salmon, the apparent 

 thermal preferendum was 3° to 5°C wide and 

 changed seasonally. Manzer et al. (1965) reported 

 that most sockeye were caught by research gill 

 nets in the North Pacific Ocean and Bering Sea in 

 waters of 4° to 6°C in May, 4° to 7°C in June, 8° 

 to 12°C in July, and 9° to 12°C in August. Based 

 on the occurrence of sockeye salmon in large 

 areas of the North Pacific Ocean, French and 

 Bakkala (1974) concluded that they are not exclu- 

 sively associated with specific oceanic conditions. 



To determine the ways in which temperature 

 might directly affect sockeye salmon, we must 

 ascertain the horizontal and vertical distribution 

 of temperatures which they experience at sea on 

 their homeward journey. An oceanographic sur- 

 vey of the North Pacific Ocean and Gulf of Alaska 

 from 16 to 24 July 1959 (S.I.O.U.C. 1965) provide 

 useful data to suit this purpose (Fig. 8). Tempera- 

 tures at 0, 30, and 50 m depth were used to esti- 

 mate the extent of horizontal and vertical gradi- 

 ents experienced by salmon migrating to the 

 northern tip of Vancouver Island along the path 

 which Fraser River sockeye seem to take. 



If we assume that salmon swam 48 km/day on 

 the surface along the route of the ship, they would 

 have experienced total temperature changes of 

 + 3.68°C or a daily average of +0.10°C/day (Table 

 3). Averaged over the stations, vertical excur- 

 sions from to 30 m would have caused the fish to 

 experience changes of -0.81°C. Dives from the 

 surface to 50 m would have been accompanied by 

 changes averaging -5.00°C. Earlier in the sum- 



464 



