136 



Fishery Bulletin 90(1). 1992 



pollock spawning seems to be timed and positioned to 

 correspond to this influx. In years when this influx is 

 delayed or absent, survival of larvae may be reduced 

 (Nakatani and Maeda 1989). Years with an early inva- 

 sion of the Oyashio Water have resulted in large year- 

 classes of walleye pollock (Nakatani 1988). However, 

 a strong year-class was also observed in 1980 when 

 there was a late invasion (Nakatani and Maeda 1983, 

 Nakayama et al. 1987). To predict population size fluc- 

 tuations will require further studies on the causes of 

 larval mortality. 



Besides factors influencing larval food production in 

 Shelikof Strait (Incze et al. 1990), the complex dynam- 

 ics of the ACC as it exits the strait seem important in 

 determining the rate of drift of the larval patch and 

 its resultant position when the larvae are ready to settle 

 (Reed et al. 1989). If the larvae are in the center of the 

 ACC as it exits the strait, they may be carried quickly 

 offshore through the sea valley between the Semidi 

 Islands and Chirikof Island, as apparently happened 

 in 1985 (Incze et al. 1989). Some of these larvae may 

 remain offshore where larval feeding conditions are 

 probably not ideal. The return of offshore larvae to the 

 shelf for demersal settlement is also problematical. If 

 the larvae are on the Alaska Peninsula side of the core 

 of the ACC as it exits the strait, their drift will be 

 slower, and they should remain in the coastal region 

 where food production is probably enhanced. Their tra- 

 jectory should carry them west along the Alaska Penin- 

 sula to shelf areas suitable for demersal settlement. 



Storm winds blowing offshore from Wide Bay may 

 displace the ACC as it exits the strait, and eddies have 

 been observed in this area. The influence of such fac- 

 tors on the larval patch and larval food production may 

 be important in determining the numbers of larvae 

 reaching the juvenile stage. 



Conclusions 



It appears that within large areas of distribution, wall- 

 eye pollock populations have evolved to spawn in very 

 specific areas and during brief times of the year. Adults 

 migrate to these areas annually for spawning. This 

 spawning pattern produces concentrations of plank- 

 tonic eggs and larvae that far exceed those reported 

 for any other fish (> 20,000 eggs/m^; <5000 larvae/ 

 m^). These spawnings are such that the eggs and 

 larvae find themselves in areas where suitable food is 

 abundant and where currents later carry larvae to 

 suitable nursery areas. It appears that interannual 

 variations in oceanographic conditions responsible for 

 food production and larval drift impact larval survival, 

 and hence year-class strength. Although there are 

 marked differences in the geography and oceanography 



of Shelikof Strait and Funka Bay, walleye pollock have 

 adapted to reproduce successfully in both areas. Adap- 

 tations in the early life history of walleye pollock to 

 these differences in environment include timing and 

 duration of the spawning season, specific gravity of the 

 eggs, and differences in prey size in relation to larval 

 size. 



Time of spawning in both areas corresponds to sea- 

 sonal transitions in hydrographic conditions (Nakatani 

 1988, Kim 1987). The spawning season is several 

 months long in the lower-latitude Funka Bay area 

 where there is considerable interannual variation in 

 timing of the intrusion of the cold Oyashio Water, 

 which increases copepod naupliar production. The 

 Shelikof Strait area spawning is very peaked, taking 

 place mainly over a few weeks and during the same 

 time each year, early April. This is the time when 

 currents are at an annual minimum due to reduced 

 precipitation and weak winds. We do not know if low 

 current strength is the seasonal signal that fish respond 

 to, but presumably the signal is less variable than the 

 intrusion of Oyashio Water. 



Eggs are less dense in Funka Bay where water 

 depths are only about one-third those of Shelikof Strait. 

 In Funka Bay, the eggs rise in the water column after 

 spawning and drift into the inner part of the bay. In 

 Shelikof Strait, the eggs remain in the nearbottom 

 water where they are spawned and show no appreciable 

 drift. This difference in transport of eggs may relate 

 to the desired location of hatching. Copepod produc- 

 tion is enhanced when Oyashio Water enters Funka 

 Bay and the egg drift pattern enables the eggs to hatch 

 there. In Shelikof Strait, the upper layers of water dur- 

 ing the spawning season are moving to the southwest 

 at a rate that would flush eggs in surface waters out 

 of the strait and into the offshore Alaska Stream in 

 a few weeks. By remaining in the sluggish bottom 

 waters, hatching is more likely to occur in southwest 

 Shelikof Strait where larval prey may be more abun- 

 dant. Interannual variations in storms in this area may 

 effect copepod production and thereby larval condition. 



In both areas, nauplii of species of small copepods, 

 Pseudocalanus and Oithona, are dominant in the diet 

 of first-feeding larvae. Eating small prey is energetical- 

 ly costly for larger larvae, so it may be critical for them 

 to encounter more advanced stages of copepods (Incze 

 et al. 1984). This may be more important in Shelikof 

 Strait than in Funka Bay because larvae in Funka Bay 

 start eating larger prey at a smaller size than do larvae 

 in Shelikof Strait. 



Drift of larvae to nursery grounds is more important 

 in Shelikof Strait than it is in Funka Bay. It appears 

 that most juveniles that result from spawning in Sheli- 

 kof Strait inhabit shelf and nearshore areas 100-200 km 

 from the spawning location by the age of 4 months 



