388 



Fishery Bulletin 93(2). 1995 



during daylight. During darkness in November and 

 December, they moved into shallows of 9 m or less 

 and became more loosely aggregated; some individu- 

 als remained on the bottom and motionless under a 

 light beam. At dawn the loose aggregations became 

 shoals and moved downslope again to depths >30 m. 



Discussion 



The consistent appearance of the age-0 pollock that 

 I observed each year for 22 years at the Auke Bay 

 site agrees during 1986-89 with the earlier appear- 

 ance for those years of pollock larvae observed in 

 Auke Bay by Haldorson et al. ( 1990). They found that 

 little hatching occurred after the major larval cohort 

 appeared and concluded that pollock spawn during the 

 relatively short period when larval feeding conditions 

 are optimal. 



By July, I observed that the larvae had usually 

 settled out from a planktonic existence and had be- 

 come demersal juveniles. Typically at this time, sea 

 surface temperatures in Auke Bay are near a yearly 

 peak (Bruce et al., 1977), a thermocline has formed 

 around 12 m depth, zooplankton abundance in the 

 upper water column approaches its yearly maximum 

 (Carlson, 1980), and 3—4 cm juvenile pollock feed 

 mostly on the copepod Acartia clausi (Krieger, 1985). 

 Similar to my findings, the sculpin Myoxocephalus 

 sp. is a major predator of age-0 juvenile pollock in 

 the Bering Sea (Smith, 1981). 



By October, when the young pollock were 8-10 cm 

 TL, they appeared sufficiently motile to evade most 

 bottom-dwelling predators. They moved to deeper 

 water in the fall around the time that, typically, 

 storms mix the surface and mid-water strata, the 

 thermocline breaks down, and sea temperatures de- 

 crease sharply and become nearly uniform through- 



