146 



Fishery Bulletin 88(1), 1990 



cruise permits a comparison among years, and provides 

 an estimate of the increase in length of fish in the 

 population with time (Fig. 9). The mean length of larvae 

 in the population at a particular time results from a 

 combination of time since spawning and individual 

 growth, as well as the effects of mortality. Older lar- 

 vae will have been subjected to mortality for longer 

 periods than younger larvae, so the change in length 

 of larvae in the population should be less than individual 

 growth rate. In 1978, 1979, and 1983, only one cruise 

 sampled larvae. However, in 1981 and 1985, five and 

 four cruises, respectively, sampled larvae. Altogether, 

 20 cruises were available to assess larval-population 

 length increase. 



Before the third time interval, little increase in the 

 length of larvae in the population or variation among 

 years in size of larvae was evident. The mean lengths 

 were 3-6 mm. Larvae at the end of April were smallest 

 in 1981 and largest in 1985. After the middle of May, 

 the larval population length seemed to increase at a 

 fairly uniform and linear rate through the end of May. 

 In 1981, 1982, and 1985, larvae in late May were be- 

 tween 7 and 9 mm. In 1983, however, larvae were over 

 10 mm in late May. Excluding the larval lengths from 

 1983, the length data from the third through fifth time 

 intervals fit the straight line of 



length (mm SL) = -9.60 + 0.122  day of year. 



1 1 1 1 1 1 1 1 1 



-B- 32 

 -E- 86 



Figure 9 



Lengths of walleye pollock larvae l)y .sanipling date and year. 



with r- = 89.7. Thus the population increased in 

 length at a rate of about 0.12 mm/day. Based on 

 analysis of larval otolith daily growth increments, in- 

 dividual growth in 1983 was 0.2 mm/day (Kendall et al. 

 1987). 



Distribution of cohorts of eggs and larvae 

 in Shelikof Strait 



The five time intervals of 16-18 days were about equal 

 to the incubation period of the eggs, and the larvae 

 grew about 3 mm (~0. 1 7 mm/day) during each time in- 

 terval, after hatching at 3-4 mm. Thus eggs from one 

 time interval were assumed to be the 3-6 mm larvae 

 observed in the following time interval, the 6-9 mm 

 larvae in the time interval after that, and so on. Under 

 this assumption three cohorts (1-3) were established, 

 based on eggs from the first, second, and third time 

 intervals, regardless of year, in order to examine the 

 geographic displacement of the eggs and larvae dur- 

 ing development (Table 6). Changes in abundance with 

 time in the cohorts is influenced by differences among 

 years in sampling and overall population abundances, 

 which prevent estimation of mortality rates. 



Eggs in the first cohort, those from the first time in- 

 terval, were slightly northeast of larvae in this cohort, 



which were represented by 3-6 mm larvae from the 

 second time interval (Fig. 10). The centroid of larvae 

 that were 3-6 mm long in the second time interval was 

 in the southern end of Shelikof Strait, midway between 

 the southern end of Kodiak Island and the Alaska 

 Peninsula. In the third time interval the first cohort 

 (6-9 mm larvae) was centered further to the southwest, 

 just north of a line between Sutwik and Chirikof 

 Islands. In tlie fourth time interval the 9-12 mm lar- 

 vae were centered near the position of the centroid of 

 the 3-6 mm larvae in the second time interval. The 

 ellipse for these larvae was quite elongate (Table 6), 

 indicating considerable dispersion on the along-strait 

 axis. This pattern reflects differences in the distribu- 

 tion of this size larvae among years, particularly in the 

 abundance of these larvae in the northern part of 

 Shelikof Strait in 1985. The centroid of the 1 2- 1 5 mm 

 larvae from the fifth time interval is positioned in the 

 vSemidi Islands, and the ellipse is broadened in the 

 across-strait dimension compared with the previous 

 ones. Thus, excluding the fourth time interval, which 

 was influenced by the aberrant pattern of 1985, the 

 first cohort of eggs and larvae appears to have been 

 advected to the southwest in the center of the strait 

 from the southern end of Kodiak Island to the Semidi 

 Islands between early April and late May. 



