Kendall and Picquelle Egg and larval distributions of Thersgrs chslcogramma 



145 



between Sutwik and Chirikof Islands had a few larvae. 

 During the third time interval most sectors in Shelikof 

 Strait proper contained large numbers of larvae. The 

 larvae seemed to be most abundant along the Alaska 

 Peninsula. South of Sutwik Island, most of the sectors 

 contained a few larvae, with most larvae tending to oc- 

 cur in sectors along the deep trough and along the con- 

 tinental shelf margin. In the fourth time interval lar- 

 vae were mainly in lower Shelikof Strait south to 

 Chirikof Island. They were more abundant in the center 

 of the strait over the deep trough than along the Alaska 

 Peninsula. Larvae were present in nearly all sectors 

 during the fifth time interval, and they were more 

 widespread and relatively more abundant south of Sut- 

 wik Island than previously. Sectors with highest con- 

 centrations of larvae were between the south end of 

 Kodiak Island and Chirikof Island. 



Centroids of the larval distributions liy time interval 

 represent the results of spawning locations and advec- 

 tion and mortality of the eggs and larvae (Fig. 7). Dur- 

 ing the second time interval, the centroid of the larval 

 distribution was over the deep trough between the 

 Trinity Islands and the Alaska Peninsula, slightly south 

 of the centroid of eggs during the previous time inter- 

 val. The location of this larval centroid probably largely 

 reflects advection of larvae from the eggs sampled in 

 the previous time interval. These few larvae are not 

 apparent in later time intervals, probably because their 

 numbers are overwhelmed by lai-vae produced from the 

 more intense later spawning near Cape Kekurnoi. The 

 centroid of larvae during the third time interval was 

 in the center of the strait between Cape Kekurnoi and 

 Wide Bay. This is north of the larval centroid, but 

 southwest of the egg centroid, from the previous time 

 interval. Samples from the second time interval ac- 

 counted for most of the eggs, whereas samples from 

 the third time interval accounted for most of the larvae. 

 The larvae collected in the third time interval primar- 

 ily resulted from the eggs observed in the second time 

 interval. The difference in location of the egg centroid 

 from the second time interval and the larval centroid 

 from the third time interval is probably largely due to 

 advection to the southwest of eggs and larvae between 

 these intervals. The larval centroid from the fourth 

 time interval was displaced slightly to the south of the 

 larval centroid from the third time intei-val. The sizes of 

 the ellipses from these two time intervals were essen- 

 tially equal, indicating that physical diffusive processes 

 were balanced by biological processes of recruitment 

 and mortality and physical processes counteracting 

 dispersal. The larval centroid from the fifth time inter- 

 val was displaced to the southwest from that of the 

 fourth time interval, and the ellipse was considerably 

 larger. The shape of the ellipse indicated that the larvae 

 were more dispersed in both the along- and across- 



I I I 



I I 



- 56 00 



- 55 00 



Figure 8 



Centroids of walleye pollock larval distrilnitions in Shelikof Strait 

 during the fifth time interval (16 May-2 June). 1979-86. 



strait directions than previously. Based on the low num- 

 bers of eggs collected after the second time interval, 

 the larval centroids from the fourth and fifth time in- 

 tervals largely reflect displacement of eggs and larvae 

 first seen as eggs during the second time interval. 



Interval 5 by year The fifth time interval evinced 

 considerable variation in the location of the larval cen- 

 troids among the years sampled (Fig. 8). The centroid 

 of larvae in 1985 was off the westernmost coast of 

 Kodiak Island. The ellipse from 1985 was quite elongate 

 in its along-strait axis due to the occurrence of larvae 

 further north in Shelikof Strait than seen in any other 

 year sampled. The centroids from 1979, 1981, and 1986 

 were close to one another over the trough between Sut- 

 wik and Chirikof Islands. In 1983, the centroid was in 

 the Semidi Islands, and in 1982 it was further south, 

 between Chirikof Island and the Alaska Peninsula. The 

 centroid from 1985 was 290 km northeast of the cen- 

 troid from 1982. 



Larval lengths A plot of weighted mean length of lar- 

 vae versus weighted mean sampling date* for each 



'Weighted mean sampling date = D. where 

 J-N,  D, 



N^ = number/10 nr in sample i, and 



£>, = day of year when sample i was taken. 



