550 



Fishery Bulletin 92(3), 1994 



May. Although many factors probably 

 contributed to the distribution of 

 eggs, their distribution was positively 

 related to temperature, which in- 

 creased with depth. 



Kendall and Kim ( 1989) developed 

 a model, based on field collections 

 and laboratory experiments, to de- 

 scribe the vertical distribution of 

 walleye pollock eggs from Shelikof 

 Strait in relation to water density. 

 One of the model's assumptions was 

 that the specific gravity of eggs does 

 not vary interannually. Based on this 

 assumption and their observations of 

 the changes in egg buoyancy and 

 depth distribution during develop- 

 ment, eggs would rise to different 

 depths in the middle stages of devel- 

 opment depending on water density 

 in particular years. The vertical dis- 

 tribution might then influence the 

 horizontal distribution of eggs if 

 there was a significant vertical shear in the water 

 column. Like Kendall and Kim (1989), we found 

 middle-stage pollock eggs at shallower depths than 

 early or late stage eggs. Our data, however, suggest 

 that the depth distribution of eggs changes during 

 development regardless of water density. Pollock egg 

 density appears to vary interannually — in the eight 

 tows from which significant numbers of eggs were 

 collected and for which concurrent hydrographic data 

 are available, the mean depth of occurrence varied 

 from 153 to 206 m, and the density varied from 25.31 

 o t to 25.63 o r Among the four years for which Kendall 

 and Kim (1989) present data (1977, 1981, 1985, 

 1986), density in the middle layer of the water col- 

 umn ( 162-216 m) varied interannually from 25.58 a, 

 to 25.87 a r and modelled depth distribution of 

 middle-stage eggs varied from about 160 to 230 m. 

 While the ranges of density and depth of eggs seen 

 in the present study are similar to those modelled 

 by Kendall and Kim (1989), the proposed relation- 

 ship between depth of eggs and water density is not 

 evident. A relationship might have been seen if a 

 greater range of water densities had been found. The 

 temperature and density of the water in which most 

 of the eggs were found in the present study ( 1987 

 and 1988) most closely resembled the values reported 

 for 1981 by Kendall and Kim (1989). 



Ingraham et al. ( 1991 ) compared long-term annual 

 means of water temperature, salinity, and density 

 at 225 m depth in Shelikof Strait with values in in- 

 dividual years when circulation in the Gulf of Alaska 

 was anomalous. They found high values for all three 



24 36 48 60 72 



Time (hours since 0000 22 May) 



84 



'J 6 



— * — Wind speed 



Larval mean depth 



Night 



Figure 10 



Mean depth of walleye pollock, Theragra chalcogramma, larvae and 

 wind speed during series nine. 



variables in 1985; this was a year when water on the 

 continental slope did not include fresher, colder wa- 

 ters from the eastern Gulf of Alaska owing to reduced 

 westward transport. Of the years considered here 

 (1985-88), only 1985 was characterized by anoma- 

 lous flow conditions that could have produced un- 

 usually warm, dense (>5.4°C, >26.2 a,) bottom water 

 in Shelikof Strait (Ingraham et al., 1991). According 

 to the model in Kendall and Kim (1989), the eggs 

 should have risen closest to the surface in 1985. How- 

 ever, mean depth of eggs in the four tows in 1985 

 was 220 m, in the five tows in 1986 it was 211 m 

 (Kendall and Kim, 1989), in the three tows of series 

 one in 1987 it was 200 m, and in the seven tows of 

 series two in 1988 it was 176 m. 



In series three through eight, the mean depth of 

 occurrence of eggs was less than 130 m. Given the 

 low numbers of eggs and the sampling intervals de- 

 signed mainly to sample larvae, these values are not 

 robust. Kendall and Kim ( 1989) also found some eggs 

 with significantly lower density than others in their 

 specific gravity experiments. The data presented here 

 confirm that some eggs have a low specific gravity 

 value, but that these are infrequent and occur pri- 

 marily later in the season after the majority of eggs 

 have hatched. 



Apparently after hatching, larvae move quickly to 

 the upper part of the water column. Both eggs and 

 larvae in series two had mean depths of occurrence 

 between 153 and 212 m among the seven tows. The 

 mean length of larvae in series two was 3.8 mm, 

 which is within the range of size at hatching (Kendall 



